TL;DR

• Alexander Marcus, CEO of ESI Total Fuel Management, is a speaker at the free online Kingdom Impact Summit, Nov. 16-20, 2026. IBAM’s Steve Adams hosts it.
• The summit is for Christian entrepreneurs, business owners, pastors, ministry leaders and marketplace professionals.
• The theme is “One Village Can Spark a Global Kingdom Movement.” More than 30 speakers are taking part, including Jordan Raynor, Mark Batterson, Peter Greer and Dr. Ché Ahn.
• New sessions are released daily, and each day’s sessions are free to watch for 24 hours. Closed captions are included.

# # #

By Alexander Marcus, CEO, ESI Total Fuel Management

I’m speaking at the free online Kingdom Impact Summit, Nov. 16-20, 2026. Join me for practical conversations on faith, work, and serving others.

What would change if your faith shaped the way you work, lead, and use what God has given you?

I’ve spent more than 40 years in engineering, fuels, and critical power, and 30 of them leading ESI Total Fuel Management. Along the way, I’ve come to believe that business is a stewardship: of people, resources, communities, and the environment. That’s why I’m glad to be a speaker at the Kingdom Impact Summit 2026. It’s a free online event on Nov. 16-20, hosted by Steve Adams, founder and president of IBAM.

Save your free place: summit2.ibam.org/sessions.

Who it’s for

The summit is for Christian entrepreneurs, business owners, pastors, ministry leaders, and marketplace professionals. It’s also for anyone who wants their work and resources to matter beyond the paycheck. You don’t need to know anything about business as a mission. You only need to be willing to take one faithful next step.

What to expect

The theme is “One Village Can Spark a Global Kingdom Movement.” Each day has its own focus:

• Day 1: See your work through God’s eyes
• Day 2: Lead with wisdom and steward for impact
• Day 3: Make your values visible at work
• Day 4: Use success to serve others
• Day 5: Put your gifts to work, here and around the world

More than 30 speakers are taking part, including Jordan Raynor, Mark Batterson, Peter Greer, and Dr. Ché Ahn.

How to watch

New sessions are released each day from Nov. 16-20. Each day’s sessions are free to watch for 24 hours after release, and closed captions are included. You don’t have to watch everything. Pick the conversations that fit the questions you’re carrying right now.

One thing to try first

Ask yourself: who could I serve better through the work I’m already doing? Choose one person and one action this week. That is often where faith at work begins.

Register for free here: summit2.ibam.org/?am_id=alexandermarcus

The summit is hosted by IBAM, a faith-driven organization that equips entrepreneurs in underserved communities with startup support, business training, and discipleship. After you register, you’ll see an optional All-Access Pass for people who want to keep the sessions.

The post What Would Change If Your Faith Shaped Your Work? Join Me at the Kingdom Impact Summit 2026 appeared first on Data Center POST.

TL;DR

• Mexico’s digital infrastructure market is expanding beyond established hubs such as Querétaro, creating new opportunities for data center development and regional investment.
• Public-private partnerships are playing an increasingly important role in addressing power availability, permitting, connectivity and infrastructure financing.
• New fiber routes, Internet Exchange Points (IXPs) and regional network operators are critical to strengthening Mexico’s connectivity ecosystem and cross-border digital corridors.
• Evolving telecommunications regulations and initiatives to expand rural connectivity are influencing the country’s long-term digital infrastructure strategy.

# # #

Mexico’s emergence as a strategic market for data centers, artificial intelligence (AI) and connectivity is opening new opportunities for infrastructure development across the country. While established markets continue to attract investment, emerging regions, evolving telecommunications policies, and growing demand for cross-border connectivity are changing how Mexico approaches its digital future.

These priorities shaped the agenda at Mexico Connect 2026, held September 9–10 in Mexico City. The two-day event explored the investments, partnerships, and infrastructure requirements supporting Mexico’s expanding digital economy, with discussions spanning hyperscale development, energy availability, fiber networks, regulatory reform, and digital inclusion.

Data Center Development Moves Beyond Traditional Markets

As demand for hyperscale and AI infrastructure increases, Mexico’s data center industry is exploring opportunities beyond Querétaro. Regions including Nuevo León, Jalisco, Hidalgo, San Luis Potosí, and Yucatán are attracting attention as potential locations for new facilities, supported by regional development initiatives and evolving infrastructure requirements.

Discussions examined the factors influencing site selection, including power availability, land, permitting, workforce readiness, and connectivity. The importance of coordination among developers, utilities, and regional governments was another focus, particularly as projects move into markets without established data center ecosystems.

These emerging opportunities also highlight the role of public-private partnerships in accelerating development. Aligning infrastructure investment with energy planning, regulatory support, and regional economic priorities will be essential to expanding Mexico’s data center footprint while supporting long-term growth.

Fiber Investment Strengthens Mexico’s Regional Connectivity

Mexico’s growing digital economy requires a connectivity ecosystem capable of supporting hyperscale facilities, cloud services, and increasing data traffic. Investment in long-haul fiber networks, Internet Exchange Points (IXPs), and Internet Service Providers (ISPs) is becoming increasingly important to improving network performance and expanding access to digital services.

The development of new cross-border fiber corridors was a major topic, particularly projects connecting Mexican infrastructure hubs with markets in the United States. Proposed and developing routes, including connections between Querétaro and Phoenix, demonstrate the opportunities to strengthen Mexico’s position as a digital bridge between North and Latin America.

Beyond international connectivity, discussions addressed the importance of improving domestic traffic exchange, reducing latency, and strengthening regional networks. As demand for AI and cloud services grows, efficient interconnection and a more extensive fiber backbone will be critical to supporting Mexico’s digital infrastructure expansion.

Regulatory Changes and Financing Influence Infrastructure Growth

Mexico’s evolving telecommunications regulatory environment is introducing new considerations for operators, investors, and infrastructure developers. The transition to the Comisión Reguladora de Telecomunicaciones (CRT) has raised questions surrounding spectrum policy, market competition, and future investment priorities.

The agenda also examined how financing models are adapting to support large-scale infrastructure projects. Development finance institutions, commercial banks, private investors, and international capital providers are playing important roles in funding fiber networks, backbone expansion, and connectivity initiatives.

As Mexico pursues additional infrastructure investment, regulatory clarity, and access to financing will be important factors in determining project viability. Greater alignment between public policy and private capital could help accelerate deployment while supporting competition and broader access to digital services.

Expanding Connectivity Beyond Major Urban Centers

While large-scale data center projects and international fiber corridors are important to Mexico’s digital economy, expanding connectivity to underserved communities remains a national priority. Smaller and regional telecommunications operators have an important role in reaching areas where traditional network deployment models may be less economically viable.

Discussions explored policy incentives, infrastructure sharing, and financing mechanisms designed to support these operators. Initiatives involving PROMTEL and other public-sector programs were also examined as potential ways to strengthen rural broadband deployment and improve last-mile connectivity.

These efforts emphasize the need for a more inclusive approach to infrastructure development. Supporting regional operators and encouraging collaboration between government and industry can help extend the benefits of digital investment beyond established commercial and technology centers.

Mexico’s Next Phase of Digital Growth

Mexico Connect 2026 highlighted the opportunities emerging as the country strengthens its position within the North American digital infrastructure ecosystem. From expanding data center development into new regions to improving cross-border connectivity and supporting underserved communities, the agenda reflected the importance of coordinated investment across multiple infrastructure sectors.

Mexico’s continued growth will depend on how effectively industry stakeholders address energy requirements, financing, regulatory changes, and network expansion. With public and private organizations working to align these priorities, the country has an opportunity to strengthen its connectivity infrastructure and support a more geographically diverse digital economy.

To explore more of the trends, opportunities, and developments influencing the global digital infrastructure industry, visit Capacity Media’s upcoming events calendar and discover where industry leaders will gather next.

The post Beyond Querétaro, Mexico’s Digital Infrastructure Opportunity Expands Nationwide appeared first on Data Center POST.

TL;DR

• Low public awareness and narrow misconceptions create a major hiring bottleneck for the data center industry as it struggles to build a talent pipeline to keep pace with exploding demand.
• Academic degree requirements obscure practical needs, turning away school leavers and tradespeople even though facility operations rely heavily on mechanical expertise and traditional trade disciplines.
• Recruitment must leverage transferable skills by drawing candidates from adjacent technical trades like plumbing, HVAC, and electrical work, as well as non-technical fields like hospitality and logistics.
• Bridging the workforce gap requires demystifying facilities through direct school outreach like the Data Centre STEM Ambassadors Programme and establishing modular, hands-on training pathways

# # #

By Stuart Newman, Business Development Manager at Apx Data Centre Solutions

The data centre sector forms the invisible backbone of modern life: the phone you’re holding, the video you’re streaming on the commute to work, every financial transaction, AI query and cloud-stored document. It all relies on complex physical infrastructure operating 24 hours a day, 365 days a year.

In spite of the role they play in our ‘always-on’ digital economy, the general understanding of the infrastructure inside these windowless warehouses remains a mystery. But it’s a predicament with severe consequences: the next generation simply doesn’t know we exist or what opportunities lie inside those walls. And that lack of awareness is having a major knock-on effect, with the industry facing an increasingly urgent challenge to secure the talent pipeline to keep pace with the exploding demand.

With this year’s National Data Centre Day focusing on the theme of ‘education’, it provides us, as industry professionals, with an opportunity to focus on how we recruit, train and demystify the sector for the next-generation. Because right now, our biggest bottleneck isn’t the pace at which technology is changing. It’s perception and access.

The misconception of technology

You only have to take a cursory glance at our news outlets today to see just how people feel about data centres. Around the globe, they’re lambasted for being monolithic, featureless warehouses filled with complicated computer racks.

Because of the picture that’s (rightly or wrongly) being painted, there’s a widespread assumption that working within one requires a specialised computer science degree or an elite background in software engineering.

It’s a narrow perception, often perpetuated by the mysterious aura that tends to hang over data centres,  that turns away bright young school leavers and skilled tradespeople before they’ve even considered applying.

At the same time, traditional educational pathways and standard apprenticeship frameworks often struggle to keep pace with the relentless speed of innovation in critical infrastructure. The result? A system where newcomers see only blockers rather than entry points, leaving a gap precisely where structured, hands-on training should be.

By focusing primarily on formal academic credentials, we obscure a fundamental truth: the digital world is not built solely on lines of code. Instead, it’s built on practical problem solving, high spec mechanical expertise and, crucially, on traditional trade disciplines.

In fact, we’re starting to see the latter play an important role in shoring up our industry. A recent Data Centre Dynamics Workforce report reaffirms this, by noting that organisations in the sector are “increasingly looking to adjacent technical infrastructure and utilities sectors to fill roles”.

And for me, this is something I’ve seen first-hand.

Early interest

While education-based pathways remain an important route to driving interest in the next generation of data centre professionals, more than a quarter of DCD’s respondents cited low industry awareness as a blocker to entry.

You only have to look at the official National Data Centre Day’s recent social media campaign to know that we’ve got an awareness issue. The series, which marked the industry’s second anniversary as National Critical Infrastructure, was called ‘then and now’ and aimed to document the realities of what school-age children aspire to become as adults.

Throughout the campaign, we’d seen posts about aspiring primary school teachers, pop stars and theme park enthusiasts, engineers, soldiers and athletes, among others. I, myself, wanted to be a motorsport engineer.

Besides it being a great tool to show the range of roles those individuals now hold in the industry, it’s a campaign that’s demonstrated two things: the first, that none of them knew this industry, or the roles within it, existed at a younger age. The second, just how many transferable skills exist across other industries and how interests can crossover.

All of this begs the question: without awareness at school level, how can we expect young minds to take an early interest? It starts, I think, with operators and suppliers being more open-minded; being willing to open doors to candidates with non-linear backgrounds and alternative skill sets.

The power of transferable skills

Take my own route into the industry. Having started my career as a plumber before transitioning into commercial sales and business development within the data centre sector, I’ve seen first-hand how directly these core trade skills crossover into high-tech environments.

A modern data centre is, at its heart, an exceptionally complex engineering site. Keeping mission-critical servers online requires sophisticated liquid and air cooling networks, precise environmental controls, high-voltage power distribution and continuous physical maintenance.

Skilled tradespeople such as electricians, HVAC technicians, pipefitters, plumbers and mechanical engineers already possess the core competencies these facilities need to function smoothly. So much so, it’s no surprise to see that, according to the DCD Report, as much as 19% of the workforce is being pulled from ‘expanded industries’ like utilities.

They understand pressure systems, fluid dynamics, fault diagnosis and strict safety compliance. Yet, despite possessing the exact real-world experience our industry desperately craves, many skilled workers rule themselves out of the running simply because they don’t realise how applicable their abilities are.

Beyond technical knowledge, non-technical roles across site operations, client account management, compliance and project logistics rely heavily on soft skills that are honed in entirely different sectors: think hospitality managers who excel under pressure and deliver exceptional client service, aviation cabin crew trained in rigorous safety protocols and crisis management, a logistics coordinator who wrangles complex schedules and inventory control or, perhaps, an event planner who knows how to manage vendor ecosystems and strict deadlines.

The reality though, is that none of these professionals are likely to even consider a data centre as their next logistical step, despite their core strengths being exactly what today’s modern facility management and data centre teams need.

This is precisely why we must challenge how we view candidate pools. Right now, according to the DCD Report, around 29% of the workforce is currently drawn from critical infrastructure. And this is where we risk pigeon-holing ourselves; when we look too closely at sectors that mirror our own, we aren’t creating a broader talent pipeline. We’re simply shuffling people within a very similar ecosystem.

To build genuine, long-term resilience, we must look beyond obvious structural parallels.

Redefining education for the infrastructure era

If we’re to bridge the growing skills gap, the industry can’t afford to wait for conventional education models to catch up. We must take an active role in demonstrating that university isn’t the only gateway to high-growth careers.

To truly inspire the next-generation, I think three fundamental changes need to happen.

We must demystify the facilities themselves. Industry leaders must open their doors to schools, colleges and trade academies to show students the tangible reality of these modern facilities. We’re starting to see this happen in pockets around the industry, primarily in an effort to bring down the proverbial walls and slowly break down the growing, negative perception building around the market.

In fact, just the other day, I was reading a story about DataVita boss, Danny Quinn, inviting Greens co-leader, Ross Greer and Dr Kat Jones, director of Action to Protect Rural Scotland (APRS), down to his facility in North Lanarkshire to help start that ‘demystification’ process and answer directly some of the criticism we’re seeing.

We’ve also got companies like Telehouse, a leading global data centre service provider, tapping into pop culture to bridge the gulf in public knowledge and tackle common misconceptions by creating an animated avatar character named ‘DC’.

But demystification also means getting out from behind our perimeter fences and into schools. A fantastic step in this direction is the newly-launched Data Centre STEM Ambassadors Programme, an initiative established by National Data Centre Day alongside AVK and STEM Learning. By connecting industry professionals directly with students aged 8 to 16 through classroom talks and mentoring, it gives us a structured route to show young minds what powering the digital world actually looks like. It’s an initiative that we, at Apx, are eager to throw our weight behind, because brokering these sorts of discussions and opening the doors, no matter the audience, is crucial to building a better future for the industry.

Once we’ve broken down the walls, we then need to look at the pathways to getting the next generation trained up. In any industry, starting from scratch can be extremely daunting. We need to put pathways in place to train new colleagues, providing them with the hands-on support needed to give them the confidence to seamlessly transition into critical infrastructure roles.

We also need modular, hands-on training initiatives that recognise existing baseline capabilities, whether from a mechanical trade, customer-facing role or emergency service background, and quickly bridge the gap to data centre application. By establishing clear, cross-skilling bootcamps and structured earn-while-you-learn pathways, we can lower the barrier to entry and show non-traditional talent that they already hold the building blocks for a successful career in critical infrastructure.

Opening the doors

The data centre sector sits at the absolute centre of global progress, but our continued growth relies entirely on the people behind the physical infrastructure.

Moving the needle requires us to be braver in how we recruit. By opening our facility doors, broadening our search parameters beyond traditional degrees and valuing transferable soft and trade skills, we can transform what was once a hidden, super specialised sector into one of the most accessible, rewarding career paths of the modern era.

If we want the next generation to take a chance on us, we have to be willing to take a calculated chance on them. It’s time to stop waiting for talent to find us and start showing them exactly where they fit in.

# # #

About the Author

Stuart Newman is the business development manager at Apx, specialising in bespoke Critical Cooling equipment engineered for both current and future operational demands.

The post Why The Data Centre Industry Must Rethink Its Talent Pipeline appeared first on Data Center POST.

TL;DR

• Parallel delivery runs two simultaneous workstreams, allowing site preparation and indoor module manufacturing to progress independently at the same time.
• Mod42’s off-site factory fabrication completes over 90% of construction indoors, eliminating weather delays, reducing on-site labor constraints, and enabling pre-testing before shipping.
• Mod42 targets deployment in under seven months by integrating complete volumetric modules with synchronized site preparation and utility commissioning

# # #

Originally posted on Mod42.

Traditional data center construction relies on a sequential approach where site grading, building construction, equipment installation, and testing occur one after another. This process takes between 24 and 36 months to complete. Since each phase depends entirely on the completion of the preceding step, unexpected delays in weather, labor, permitting, or utility interconnections can impact the entire plan. With demand for artificial intelligence and high-performance computing accelerating exponentially, this traditionally structured process becomes a major strategic liability.

Parallel delivery transforms scheduling by dividing deployment into two independent, simultaneous workstreams: the site track and the factory track. The site track manages civil work, foundations, grading, and utility connections, while the factory track produces complete data center modules indoors. With Mod42’s factory-built approach, over 90% of construction takes place in a controlled factory environment which allows projects to avoid common field disruptions such as severe weather and local labor shortages. Cooling, electrical, mechanical, and structural systems are fully integrated and tested off-site before shipping, allowing finished modules to be set into place as soon as the site pad is ready.

Successfully executing parallel deployment depends on complete volumetric modular construction, clear scope ownership, early alignment on pad dimensions and utility interfaces, and active risk communication. By coordinating factory fabrication alongside site preparation, Mod42 targets a total deployment schedule of under seven months covering fabrication, setting, integration, and final commissioning. This synchronized model drastically accelerates time-to-market compared to conventional multi-year timelines.

To read the original article, please click here.

The post How Parallel Delivery is Shortening the AI Infrastructure Deployment Schedule appeared first on Data Center POST.

TL;DR

• AI infrastructure development is accelerating, but power availability, site selection, and permitting challenges are influencing where and how new data centers are built.
• Growing demand for AI training and inference is driving innovation across compute, networking, memory, and storage infrastructure.
• Delivering AI capacity at scale requires greater coordination between infrastructure design, energy planning, capital investment, and operational strategies.
• As development expands, political considerations, regulatory requirements, and local community expectations are becoming increasingly important to project viability.

# # #

The rapid expansion of artificial intelligence (AI) is transforming digital infrastructure, creating new demands for computing capacity, power availability, networking, and data center development. These challenges were central to discussions at AI Infra Summit 2026, held September 15–17 in Santa Clara, California, where industry leaders explored the technologies, investment strategies, and operational considerations shaping the next generation of AI infrastructure.

Bringing together executives, technology innovators, infrastructure developers, and investors, the summit examined the increasingly interconnected systems required to support AI at scale. From high-performance computing and data movement to power constraints, site selection, and infrastructure economics, the discussions reflected an industry working to balance accelerated demand with the practical challenges of building and operating AI-ready environments.

Power Availability and Site Selection Are Reshaping Development Strategies

As AI workloads drive demand for larger and more energy-intensive data centers, securing viable development sites has become increasingly complex. Power availability, utility interconnection timelines, permitting requirements, and local political considerations are influencing project feasibility long before construction begins.

These challenges were central to discussions moderated by Ilissa Miller, CEO of iMiller Public Relations, who joined industry leaders examining the intersection of power, permitting, politics, and site selection. The discussion featured Raul Martynek, CEO of DataBank, and explored the factors influencing the race to secure suitable land for AI infrastructure development.

The growing importance of early-stage planning reflects a shift in how developers evaluate potential projects. Access to land and capital alone is no longer sufficient. Successful development strategies must also account for power delivery, regulatory approvals, infrastructure readiness, and the concerns of communities where facilities are proposed. These considerations are becoming essential to reducing project delays and improving long-term viability.

Scaling AI Compute Requires a More Integrated Infrastructure Approach

The increasing complexity of AI applications is placing new demands on computing architectures and the systems supporting them. As organizations expand their use of large language models, agentic AI, and real-time inference, infrastructure strategies must address performance, scalability, and efficiency across the entire technology stack.

The evolution of AI computing was explored by Ian Buck, Vice President, Hyperscale & HPC Computing at NVIDIA, whose participation in the summit included discussions on advancing infrastructure for agentic AI and the technologies supporting frontier AI development. These topics reflected the growing need to align processing capabilities with the networking, memory, and storage resources required to support increasingly sophisticated workloads.

The summit also highlighted the importance of designing infrastructure around specific AI requirements rather than relying on conventional computing models. As training and inference workloads evolve, organizations must consider how specialized hardware, distributed computing, and more efficient system architectures can improve performance while managing the cost and complexity of deployment.

Data Movement and Networking Are Critical to AI Performance

While computing capacity remains central to AI infrastructure development, the ability to move and access data efficiently is becoming equally important. Large-scale AI environments depend on high-bandwidth, low-latency connectivity between processors, memory, and storage systems, creating challenges that cannot be resolved through additional compute resources alone.

Among the experts addressing these challenges was Nick Harris, Founder and CEO of Lightmatter, who participated in discussions examining data movement, memory, and storage bottlenecks at scale. The session reflected the growing importance of improving how information moves throughout AI systems as organizations pursue greater performance and infrastructure efficiency.

Networking architecture, optical connectivity, and data transfer efficiency are increasingly important considerations for AI infrastructure investment. As distributed computing environments expand, improvements across these systems will be essential to supporting larger workloads, reducing performance limitations, and maximizing the value of computing resources.

Building AI Infrastructure for Long-Term Operational Efficiency

The rapid pace of AI infrastructure investment is creating pressure to deliver new capacity quickly while maintaining reliability, flexibility, and cost control. As facilities become larger and more technically complex, operators must evaluate how design decisions affect long-term performance, energy consumption, and the ability to adapt to future workloads.

The importance of integrated infrastructure planning was addressed by Ram Nagappan, Vice President, AI Infrastructure and Lead Architect, Gigawatt-Scale Data Centers at Oracle, whose presentation focused on the design of AI campuses at gigawatt scale. The topic highlighted the need to coordinate power infrastructure, computing requirements, and facility architecture when planning increasingly large AI deployments.

These considerations extend beyond initial construction. Operational strategies must account for changing power demands, evolving hardware requirements, and the integration of systems that support continuous, high-performance computing. Greater coordination between infrastructure developers, technology providers, and operators will be essential to managing complexity and supporting sustainable growth.

Building the Infrastructure Behind AI’s Next Phase

AI Infra Summit 2026 highlighted the scale of the challenges and opportunities facing the digital infrastructure industry. While advances in computing technology continue to accelerate AI adoption, the ability to support that growth depends on coordinated investment across power, networking, data center development, and operational systems.

The discussions reinforced that AI infrastructure cannot be planned or evaluated in isolation. From securing viable development sites to addressing data movement bottlenecks and designing gigawatt-scale campuses, the industry’s next phase will require closer alignment between technology innovation, infrastructure investment, and the communities supporting new development.

As demand for AI continues to expand, the strategies explored in Santa Clara provide a framework for addressing the practical requirements of scalable, resilient, and efficient digital infrastructure. Industry leaders will have another opportunity to continue these discussions at AI Infra Summit 10X 2027, scheduled for August 30 – September 2, 2027, in San Jose, California.

The post AI Infrastructure Growth Challenges Traditional Approaches to Data Center Development appeared first on Data Center POST.

TL;DR

• First connections: Empire Fiber Internet is now serving the first customers on Madison County’s publicly owned fiber network, which is expected to reach more than 3,600 homes, farms and businesses when completed.
• Project details and timeline: The $31 million public-private initiative spans nearly 389 miles, with construction expected to finish by early 2027. Empire is also expanding service beyond the publicly funded footprint.
• Continued momentum: The milestone builds on Empire’s recent network expansions in Ithaca, New York, and Berwick, Pennsylvania.
• Expanding access: The project expands access to connectivity that supports education, telehealth, remote work and local businesses while demonstrating how government and private providers can collaborate to address rural broadband gaps.

# # #

Empire Fiber Internet, a leading fiber optic internet service provider serving communities across New York and Pennsylvania, has begun serving the first customers on Madison County’s publicly owned fiber broadband network, marking a significant milestone in a multiyear effort to expand internet access across the rural New York county. Announced October 14, the initial connections bring the nearly 389-mile network closer to its goal of reaching more than 3,600 homes, farms and businesses, many of which previously had limited or no service.

The first serviceable portions of the network extend between the towns of Cazenovia, Sullivan, and Lenox. Additional locations will become eligible for service as crews complete construction and activate the remaining infrastructure, with construction expected to conclude by early 2027. Empire Fiber Internet will operate the network and provide internet service, while Madison County retains ownership of the infrastructure.

The $31 million Madison County Connect Project is supported by the USDA Rural Development ReConnect Program, New York State’s ConnectALL Municipal Infrastructure Program, Empire Fiber Internet, and Madison County. Its origins date to a $10.1 million USDA ReConnect grant awarded in 2021, subsequently increased to $12.7 million, to reach 1,087 unserved homes and businesses. Construction began in November 2023, and the broader project now encompasses more than 3,600 potential service locations.

Empire’s involvement also extends beyond operating the publicly funded network. The company is expanding service to additional addresses along the county’s network backbone in Cazenovia, Chittenango, and Canastota, including select neighborhoods outside the funded footprint. These extensions broaden the project’s reach while introducing another provider option for more residents and businesses.

The Madison County activation follows other recent developments for Empire Fiber Internet across New York and Pennsylvania. In August, the company expanded its Ithaca network to more than 700 additional homes and businesses, bringing its availability there to more than 5,400 locations. In September, Empire launched its fiber network in Berwick, Pennsylvania, following a $1.4 million local investment. The company also received its fourth consecutive FLX Finest Gold award for Best IT / Office Technology in 2026. Together, these milestones highlight a period of network expansion alongside continued community recognition.

With construction nearing completion, attention now turns to activating the remaining areas. Residents and businesses can check their addresses through Empire Fiber Internet’s availability tool and follow county updates as additional portions of the network open for service.

To learn more about Empire Fiber Internet, visit empirefiber.com.

The post Empire Fiber Internet Connects First Customers on Madison County’s Publicly Owned Broadband Network appeared first on Data Center POST.

TL;DR

• Standard cages and power feeds rarely fit modern enterprise workloads, so effective data center environments are now built around the customer.
• Four forces drive the shift: AI-driven density (some deployments need 50–220 kW per cabinet), mixed air- and water-cooling, diverse workloads, and data security and sovereignty requirements.
• One-size-fits-all forces a choice between overbuilding and paying for unused capacity, or underbuilding and hitting limits as demand grows. Hybrid environments across colocation, private cloud, public cloud and on-premises add fragmented vendors, contracts, and support models.
• When evaluating providers, look at power and cooling, connectivity, flexibility, compliance and support. Start with your workloads, then work backward to the infrastructure.

# # #

By Sam Zurzolo, VP of Data Center Services, 365 Data Centers

For years, data center decisions came down to a short list: how much space, how much power, and how much redundancy. Those fundamentals still matter today, but no two enterprises run the same workloads. As a result, a standard cage and a standard power feed rarely fit in most modern scenarios.

After more than three decades in telecom and data center operations, from U.S. Navy electronics to running facilities across the country at 365 Data Centers, one of the biggest changes I’ve seen is the move from standardized offerings to environments built around the customer.

What Changed

Several forces are driving the shift:

• Density: AI and data-intensive applications are pushing power requirements well past what legacy facilities were designed for. Some deployments now call for anywhere between 50 to 220 kW per cabinet.
• Cooling: Higher densities often mean air-cooled and water-cooled approaches working together, and the right mix depends on the workload.
• Workload diversity: Inference, analytics, regulated data, and traditional applications each place different demands on power, latency, and connectivity.
• Data control: Security, privacy, and sovereignty requirements increasingly determine where, and how, workloads can run.

Why One-Size-Fits-All Falls Short

A standardized environment forces compromises. Organizations either overbuild and pay for capacity they don’t use, or they underbuild and run into limits just as demand grows. Neither is efficient. Teaching fiber design early in my career taught me that the network path matters as much as the rack it connects to.

Hybrid complexity adds to the problem, because many enterprises now operate across colocation, private cloud, public cloud, and on-premises systems. Without a plan for how those pieces connect, teams end up managing fragmented vendors, contracts, and support models.

From an operations standpoint, customization also affects reliability. Power distribution, cooling design, maintenance schedules, and network paths all work best when they match the actual load. A well-matched environment is easier to maintain and easier to scale.

What to Evaluate

When assessing a provider, consider these questions:

1. Power and cooling: Can the facility support your density today, and can it grow with you?
2. Connectivity: Are carrier options, cloud on-ramps, and bandwidth available where your workloads live?
3. Flexibility: Can you start with colocation and add private cloud or other services later, rather than committing to everything at once?
4. Compliance: Does the environment meet the security standards your industry requires?
5. Support: Will experienced people work with you on design and ongoing operations, not just hand you a rack?

Start with your workloads, then work backward to the infrastructure. A clear picture of your current and projected power, latency, and compliance needs makes every other decision easier.

Planning for What’s Next

Customization doesn’t mean complexity. The goal is an environment that fits today’s requirements and can adapt as they change. Enterprises that treat infrastructure as a tailored foundation, rather than a commodity, are better positioned to deploy new technologies with confidence.

At 365 Data Centers, we work with customers to design colocation, connectivity, and cloud environments around their specific needs, with the support of a team that has operated critical infrastructure for decades. If you’re reassessing your data center strategy, we’re glad to help you think it through. Contact us today at 365datacenters.com/contact.

The post The Customization Imperative: Why Data Center Solutions Are No Longer Standardized appeared first on Data Center POST.

TL;DR

• AI-driven growth is reshaping Europe’s data center industry, placing greater demands on power availability, infrastructure design and operational flexibility.
• Battery energy storage is expanding beyond traditional backup applications to support grid stability, energy management and increasingly dynamic workloads.
• Rising compute densities are accelerating innovation in cooling technologies while challenging conventional approaches to data center design and efficiency.
• Europe’s digital infrastructure expansion increasingly depends on coordinated energy planning, connectivity investment and stronger community engagement.

# # #

As artificial intelligence (AI) accelerates demand for digital infrastructure, Europe’s data center industry faces mounting pressure to expand capacity while addressing power constraints, sustainability requirements, and increasingly complex workloads. These challenges were central to discussions at DCD Connect London 2026, held September 15-17, where industry leaders explored the technologies, strategies, and partnerships shaping the next phase of data center development.

Bringing together operators, technology providers, investors, and infrastructure experts, the event highlighted how the industry is adapting to a changing energy landscape. Discussions examined everything from battery storage and advanced cooling to AI-ready facility design, grid modernization, and the growing importance of community acceptance.

Rethinking Power Infrastructure and Energy Storage

Power availability and reliability emerged as major considerations as operators seek to support increasingly energy-intensive AI workloads. With grid connection delays and electricity costs affecting development strategies, the industry is exploring new approaches to energy storage, onsite generation, and grid participation.

The evolving role of battery storage was a key area of discussion with Liz Wallace, Senior Product Manager at ZincFive, among the industry experts examining how Battery Energy Storage Systems (BESS) are moving beyond traditional backup applications. Topics included power quality, peak shaving, load shifting, and demand response, along with opportunities to better manage fluctuating energy requirements.

As data centers become more active participants in the energy ecosystem, storage technologies may offer opportunities to improve operational flexibility and support renewable energy integration. These capabilities are becoming increasingly relevant as operators balance uptime requirements with energy costs and sustainability objectives.

Adapting Infrastructure Design for AI Workloads

The rapid evolution of AI hardware is challenging traditional data center design assumptions. Higher rack densities, changing workload profiles and closer integration between IT and operational technology are creating demand for facilities that can adapt without requiring extensive redesign.

Among the industry leaders addressing these challenges was Ivy Tavershima, Senior Technical Program Manager – AI Delivery at Microsoft, who participated in discussions examining whether standardized facility designs can remain effective as computing requirements continue to change. Topics included modular infrastructure, adaptable power, cooling systems, and the challenge of balancing repeatable designs with workload-specific requirements.

Cooling innovation was another important focus. As operators prepare for higher-density deployments, discussions explored liquid cooling, hybrid thermal architectures, and opportunities to improve efficiency without compromising reliability. The growing complexity of AI infrastructure is reinforcing the need to consider flexibility and long-term operational requirements during the earliest stages of facility planning.

Addressing Europe’s Energy and Connectivity Constraints

Europe’s digital infrastructure expansion is increasingly influenced by where sufficient power and network connectivity can be secured. Traditional data center markets face growing grid and permitting challenges, prompting developers to evaluate emerging regions with more favorable energy conditions.

The evolving relationship between data centers and energy providers was examined by industry experts including Goran Arya, Director, Energy Origination & Structuring, EMEA at Equinix. Discussions addressed grid capacity allocation, demand-response opportunities, and new models for integrating large-scale digital infrastructure into Europe’s electricity systems.

Connectivity investment must also evolve alongside changing development patterns. Joshua Seawell, Director of Product Management at Sumitomo Electric Lightwave, joined discussions exploring how fiber infrastructure, long-haul routes, and metropolitan networks must adapt as data center development expands into new geographic markets. Coordinated planning between energy providers, network operators, and developers will be essential to supporting reliable, high-capacity digital infrastructure.

Building Community Support for Sustainable Growth

Beyond technical and operational challenges, the industry is confronting increased scrutiny surrounding the environmental and economic impacts of data center development. Community concerns about energy consumption, infrastructure demands, and local benefits are influencing project planning and approval processes.

The importance of establishing a social license to operate was explored by industry representatives including Michael Blake, EMEA Community and Workforce Development, Regional Manager – Global Data Centers at Google. Discussions focused on incorporating community value into development strategies, improving public understanding, and engaging local stakeholders before opposition becomes a barrier to progress.

These considerations reinforce the importance of treating community engagement as an integral part of infrastructure development rather than a final step in the approval process. As Europe’s data center footprint expands, transparent communication and measurable community benefits will play an increasingly important role in supporting responsible growth.

Preparing for the Next Phase of Digital Infrastructure

DCD Connect London 2026 highlighted the interconnected challenges facing Europe’s data center industry. AI-driven demand is accelerating infrastructure investment, but the ability to deliver new capacity depends on more than computing technology alone. Energy availability, storage innovation, cooling efficiency, connectivity, and community acceptance are all influencing how and where facilities are developed.

The discussions underscored the need for greater collaboration across the digital infrastructure ecosystem. As operators and technology providers work to meet growing demand, flexible infrastructure strategies, and coordinated planning will remain essential to building resilient, efficient, and sustainable data centers. The industry will have another opportunity to reconnect and continue these important conversations at DCD Connect London 2027, taking place September 14-15, 2027. 

The post The Forces Reshaping Europe’s Data Center Future appeared first on Data Center POST.

TL;DR

• Ecosystem over Infrastructure: Data center value is increasingly driven by surrounding carrier density, cloud on-ramps, and Internet Exchanges rather than power and space alone.
• Compounding Business Value: Carrier-neutral environments leverage network effects to offer enhanced choice, vendor redundancy, and efficient paths for AI and hybrid cloud workloads.
• Future-Proofing Strategy: Interconnection serves as a core driver of resilient infrastructure, enabling organizations to adapt to evolving multi-cloud and distributed data needs

# # #

Written by iMiller Public Relations on behalf of 1547 Critical Systems Realty.

Modern data centers are evaluated not just by physical power and cooling capabilities, but by the strength and diversity of their surrounding connectivity ecosystems. While physical infrastructure forms the essential baseline, access to dense networks, cloud platforms, and Internet Exchanges transforms a facility into a key business driver. Organizations increasingly prioritize carrier-neutral environments where multiple providers compete, giving businesses greater freedom of choice, reduced dependence on single vendors, and improved cost flexibility.

Interconnection ecosystems operate on compounding network effects, where each added network or cloud platform creates exponential value for existing tenants and attracts further participants. This density becomes crucial for modern workloads, such as distributed artificial intelligence tasks and hybrid cloud deployments, which require direct cloud on-ramps and resilient multi-route connectivity. Consequently, facility connectivity is no longer viewed as an optional add-on, but as a central component of long-term infrastructure strategy and geographic redundancy.

Real-world deployments within 1547’s portfolio demonstrate how these ecosystem principles function in practice. At Chase Tower in McAllen, Texas, 1547 operates MCT-IX, a carrier-neutral Internet Exchange connecting over 46 networks across two meet-me rooms at speeds up to 100G, serving as a hub for cross-border traffic between the United States and Latin America. Additionally, the PTOR1 facility at Portland’s Pittock Block delivers extensive carrier access and AWS Direct Connect capabilities, allowing enterprises to link colocation environments directly to cloud platforms.

To read the original article, please click here. 

The post The Business Value of Interconnection Ecosystems appeared first on Data Center POST.

TL;DR

• Fiber operators are reassessing network expansion strategies as AI workloads increase demand for capacity, low-latency connectivity and stronger returns on infrastructure investments.
• Modular data center designs and edge computing are gaining attention as operators seek faster, more flexible ways to deploy AI-ready infrastructure.
• Community engagement and public trust are becoming increasingly important to project approvals, investment decisions and long-term development.
• Across fiber, data centers and energy infrastructure, greater coordination is essential to overcoming deployment challenges and supporting sustainable growth.

# # #

Digital infrastructure leaders gathered in Austin, Texas, September 1–3 for Metro Connect Fall 2026, where discussions focused on the opportunities and challenges shaping the future of fiber networks, data centers, energy, and infrastructure investment. Held alongside Datacloud USA and related industry forums, the event examined how accelerating artificial intelligence (AI) demand is influencing infrastructure development, financing strategies, and operational priorities.

From expanding fiber capacity and accelerating modular data center deployments to overcoming power constraints and strengthening community relationships, discussions reflected an industry navigating increasingly complex development requirements. While demand for digital infrastructure continues to grow, greater emphasis is being placed on strategic investment, operational efficiency, and collaboration across the infrastructure ecosystem.

Fiber Investment Priorities Continue to Evolve

As AI applications place greater demands on network capacity and performance, fiber operators are reevaluating how and where they invest. Timothy Parker, Chief Growth Officer at Assured Communications, joined industry leaders in examining the suitability of existing fiber infrastructure, evolving customer requirements, and the economics of new network deployments.

With investors applying greater scrutiny to capital allocation, operators face increasingly complex decisions about whether to build new networks or acquire existing assets. The growing importance of AI inference, distributed computing, and low-latency connectivity is also influencing network planning, creating opportunities for fiber providers that can align infrastructure investments with measurable customer demand.

Modular Infrastructure Supports the Growth of AI

Speed, scalability and adaptability are becoming increasingly important as data center operators respond to rising AI workloads. Valerie Crafton, Chief Revenue Officer at Mod42 and Doug Recker, CEO of Duos Edge AI, were among the industry leaders addressing how modular construction and prefabricated solutions can help accelerate infrastructure delivery.

As demand grows for both hyperscale campuses and distributed edge facilities, modular approaches offer potential advantages in deployment timelines, standardization, and flexibility. Discussions explored how prefabrication, digital design tools, and manufacturing innovations can help operators respond to changing capacity requirements while addressing construction and supply chain challenges.

Community Trust Becomes a Critical Infrastructure Priority

Public perception, local stakeholder expectations and community opposition are increasingly influencing where and how digital infrastructure projects move forward. Ilissa Miller, CEO of iMiller Public Relations, led a discussion with representatives from data center development, energy, municipal government and community affairs about the challenges of securing community support and maintaining the social license to operate.

As concerns surrounding energy consumption, environmental impacts, and local development continue to influence public opinion, developers face greater pressure to demonstrate the value of proposed projects. Discussions emphasized the importance of engaging stakeholders early, understanding community priorities and establishing credible relationships before development plans advance. These considerations are also becoming more relevant to investors as permitting delays, political risk, and community opposition can influence project timelines, market selection, and asset valuations.

Infrastructure Growth Requires Greater Industry Coordination

Across fiber, data center, energy, and investment discussions, a common challenge emerged as infrastructure development becomes increasingly interconnected. Power availability influences data center site selection, fiber connectivity determines network performance, and construction capacity affects how quickly projects can move forward.

From emerging financing structures and workforce development to behind-the-meter power strategies and network interconnection, discussions reinforced the importance of aligning infrastructure planning across traditionally separate sectors. As operators navigate more demanding technical requirements and tighter investment expectations, collaboration among developers, utilities, network providers, suppliers, and communities will play an increasingly important role.

Metro Connect Fall 2026 reflected an industry moving beyond the race to build capacity toward a more disciplined approach to growth. Meeting the demands of AI will require not only investment in technology and infrastructure, but also stronger coordination, sound financial strategies, and the public trust necessary to bring projects to completion.

As the industry continues to address these opportunities and challenges, the conversations will continue at Metro Connect 2027, February 8-10 in Hollywood, Florida. Join digital infrastructure leaders once again to exchange insights, strengthen industry connections and explore the priorities shaping the next phase of infrastructure growth. 

The post AI, Infrastructure Investment, and Community Trust Take Center Stage at Metro Connect Fall 2026 appeared first on Data Center POST.

TL;DR

• Simplified Service Delivery: Centralizing service configuration at the point of presence removes the need for repeated WAN changes, complex multicast design, and premises-level NAT adjustments.
• Scalable Multi-Service Platform: Once the initial Layer 2 segment is established, operators can easily attach new cloud, partner, or AI infrastructure services to the customer site.
• Commercial Availability: Managed by a cloud-based Path Computation Engine, the product line launches with the PBC 124-C to expand addressable markets for off-net sites

# # #

Written by iMiller Public Relations on behalf of MaiaEdge.

Connecting remote customer premises over existing IP transport allows service providers to expand their reach and service delivery without constructing new fiber infrastructure. By pairing a customer-premises device like the PBC 100 Series with a PoP-side PBC 2000, operators establish an encrypted Layer 2 SD-Ethernet extension over any available transport, including fiber, dedicated internet access, or satellite. This approach delivers a consistent Ethernet handoff directly to the customer site, bypassing the need for complex, manually negotiated network-to-network interfaces.

Traditional last-mile connections often complicate service expansion, requiring repeated WAN modifications, network address translation adjustments, and complex multicast routing configurations for every new offering. Extending an isolated Layer 2 Ethernet segment back to the operator’s point of presence solves this bottleneck by centralizing service management. Operators can seamlessly attach additional offerings, such as cloud connectivity, partner networks, or artificial intelligence infrastructure, directly at the central network level without modifying customer-premises hardware or WAN settings.

Driven by a cloud-based Path Computation Engine, the solution automates path selection and measures service performance across the encrypted connection. The initial deployment in this MaiaEdge product line, the PBC 124-C, enables operators to turn existing connectivity into an extensible platform for recurring service growth. By simplifying last-mile delivery, service providers can rapidly unlock new revenue streams and address off-net customer locations that were previously difficult to serve.

To read the original article, please click here.

The post MaiaEdge Introduces the PBC 100 Series appeared first on Data Center POST.

TL;DR

• 3C and Datalec Precision Installations have merged to form 3C Datalec, combining liquid cooling, engineering, manufacturing, connectivity, and facilities management expertise.
• The combined company forecasts more than $400 million in 2026 revenue, with approximately 1,000 employees and active projects across 16 countries.
• Backed by new investment, 3C Datalec plans to hire 200 engineers and open facilities in Poland, Canada, and the United States over the next 12 months.

###

As AI workloads increase power and cooling requirements across data centers, operators are placing greater emphasis on integrated infrastructure solutions that support projects from initial design through ongoing operations.

3C, the liquid cooling infrastructure company founded by former CoolIT Systems CEO Steve Walton, has merged with Datalec Precision Installations (DPI) to form 3C Datalec, bringing together expertise in liquid cooling, mechanical and electrical (M&E) engineering, manufacturing, connectivity, and facilities management. The combined company will provide integrated services for AI and high-density data center environments.

The newly formed organization has approximately 1,000 employees and forecasts more than $400 million in revenue for 2026. With active projects in 16 countries totaling hundreds of megawatts, 3C Datalec serves four of the five largest hyperscalers, more than 25 colocation providers, neoclouds, and enterprise customers across international markets.

Backed by new investment from Longbow Capital, S2G Investments, Apogem Capital, and QIC, the company plans to expand its global operations and technical capabilities. Over the next 12 months, 3C Datalec expects to hire 200 engineers and establish new facilities in Warsaw, Poland; Calgary, Canada; and Austin, Texas.

The company’s services will be organized around three core areas: Create, covering design and engineering; Construct, encompassing liquid cooling, M&E, manufacturing, and connectivity installation; and Care, supporting commissioning, maintenance, operations, and facilities management. This approach brings together services that are increasingly interconnected as operators address higher rack densities and more demanding cooling requirements.

“Rack densities have outgrown air, and liquid cooling is no longer optional,” said Steve Walton, CEO of 3C Datalec. “At CoolIT, we spent years preparing for this moment and became a global leader in liquid cooling. The hard part now is making cooling, power, construction and operations work as one system, while running at scale. We’re combining our liquid cooling expertise with Datalec’s engineering, manufacturing and FM capabilities, so customers have one partner from design to operation.”

Workforce development will also be a focus as the company expands. Through apprenticeships, technical training, and career development programs, 3C Datalec aims to develop additional expertise in liquid cooling and critical infrastructure while supporting the growing demand for skilled data center engineers.

The merger reflects the increasing importance of coordinated design, construction, cooling, and operational services as AI infrastructure requirements become more complex. By combining technical expertise with an established international delivery presence, 3C Datalec is expanding its ability to support both new data center developments and upgrades to existing facilities.

To learn more about the merger and 3C Datalec’s expanded capabilities, read here: www.imillerpr.com/news/3c-merges-with-datalec-to-create-us400m-revenue-liquid-cooling-and-data-centre-services-platform

The post 3C and Datalec Merge to Form $400M+ AI Data Center Infrastructure Platform appeared first on Data Center POST.

TL;DR

• AI workload demands: Rapidly accelerating rack densities and power requirements mean operators must establish accurate visibility into existing capacity before committing capital to new deployments.
• Unified operational visibility: Disconnected spreadsheets and software create visibility gaps, whereas infrastructure intelligence links facility, IT, and connectivity dependencies into a single operational model.
• Proactive planning with digital twins: Integrating operational data allows teams to simulate changes, identify power or cooling constraints early, and uncover genuine available capacity.
• Pragmatic implementation: Operators can begin by validating critical assets against physical reality, mapping system dependencies, and strengthening change management processes to support compressed AI timelines.

# # #

Artificial intelligence is pushing data center infrastructure into unfamiliar territory. Rack densities are increasing, power and cooling requirements are becoming more demanding, and AI deployments are moving at a pace that can make traditional planning cycles feel outdated.

But the challenge facing data center operators is not simply how to provide more power, more cooling, or more space. Before operators can determine what they need next, they need to know what they already have. That makes infrastructure intelligence a critical foundation for AI.

Infrastructure intelligence is the ability to maintain an accurate, connected view of the data center environment, from physical facilities and power systems to IT equipment, connectivity, dependencies, and available capacity. It gives operators the context they need to understand not only what exists, but how everything is connected, what it can support, and what could happen when something changes.

AI Raises the Cost of Inaccurate Information

In a traditional data center environment, an outdated record or incorrect equipment specification may create an operational inconvenience. In an AI environment, the consequences can be much greater. AI racks can place significantly higher demands on power and cooling than conventional workloads. A deployment that appears to have sufficient capacity on paper may encounter constraints once the actual power draw, heat load, cooling configuration, and upstream dependencies are considered.

At the same time, infrastructure information is often distributed across multiple systems, including DCIM platforms, CMDBs, facility monitoring tools, CAD drawings, spreadsheets, and other operational records. Each may contain valuable information, but they do not necessarily provide a connected picture of the environment.

The result is a fundamental visibility problem: teams may know what individual systems contain without having a reliable understanding of how those assets relate to one another. For AI infrastructure, that distinction matters.

From Asset Visibility to Infrastructure Intelligence

Knowing that a rack exists is not enough. Operators need to understand what is in the rack, how much power it actually consumes, where that power comes from, how it is cooled, what equipment it connects to, and which workloads or services depend on it. This is where infrastructure intelligence moves beyond inventory management.

By connecting facility, IT, and connectivity information, operators can establish a shared operational model of the data center. That model can help answer questions such as:

1. Can this location support another high-density AI rack?
2. What power and cooling resources would a new deployment require?
3. Which systems, services, and dependencies could be affected by a planned change?
4. Where is there genuine capacity, rather than capacity that only appears available in a disconnected system?
5. What happens if a cooling unit or power feed is taken offline for maintenance?

These are not theoretical questions; they are the questions operators need to answer before committing additional capital, deploying new infrastructure, or making changes to an increasingly complex environment.

Digital Twins Must Reflect Reality

Digital twins can play an important role in building this intelligence, but a digital twin is only useful if it remains aligned with the physical environment. A static 3D representation of a data center may provide an attractive visualization, but it does not necessarily provide the operational context required for AI infrastructure planning. A useful digital twin needs to connect the physical environment with equipment, capacity, dependencies, and operational data. 

That enables teams to move from simply seeing the environment to modeling potential changes. For example, before deploying a high-density AI rack, an operator could model its expected impact on power, cooling, and related infrastructure. Instead of discovering constraints after the hardware arrives, teams can identify them during the planning process.

Maintaining that value requires discipline. When equipment is moved, circuits are changed or infrastructure is upgraded, those changes need to be reflected in the model. Otherwise, the digital twin gradually becomes another source of outdated information rather than a trusted representation of the environment.

Infrastructure Intelligence Enables Better Decisions

The objective is not to replace every system already used by a data center operator. It is to connect the information those systems provide so teams can understand the relationships between infrastructure components. A data center infrastructure management platform such as FNT Command bridges this gap by connecting facility, IT, and connectivity data into a shared operational model. Rather than simply providing another visualization layer, it gives teams the infrastructure relationships and capacity context needed to plan deployments, understand dependencies, and make decisions based on the environment as it actually exists.

This becomes particularly important as AI accelerates the pace of change. Traditional infrastructure planning often assumes relatively predictable refresh cycles and gradual capacity growth. AI deployments can compress those timelines, requiring organizations to evaluate infrastructure decisions much faster.

The organizations that can respond effectively will not necessarily be those with the most hardware. They will be the ones that can quickly determine what their infrastructure can actually support.

Start With the Infrastructure That Matters Most

Building infrastructure intelligence does not require operators to model every asset in a data center on day one. A more pragmatic approach is to begin with the areas where AI deployments create the greatest operational and financial consequences. From there, operators can take four practical steps:

1. Reconcile records with physical reality:
   Start by validating existing infrastructure records against what is actually deployed in the data center. This creates a reliable foundation for planning and decision-making.
2. Connect assets and dependencies:
   Establish the relationships between equipment, power, cooling, and other infrastructure. Understanding these dependencies provides the context needed to evaluate the impact of a change.
3. Strengthen change processes:
   Ensure that moves, adds, changes, and upgrades are consistently captured in the infrastructure model. This helps prevent the gap between documented infrastructure and physical reality from growing over time.
4. Expand intelligence across the environment:
   Once the most consequential AI infrastructure is accurately modeled and operational processes are established, organizations can extend that intelligence across the broader data center environment.

The goal is a continuous operational discipline in which infrastructure plans are validated before changes are made, changes are reflected in the infrastructure model, and the model is routinely reconciled against reality. This creates a foundation that can scale as AI infrastructure evolves.

AI Performance Depends on More Than GPUs

The AI infrastructure conversation often focuses on processors, power availability, cooling technologies, and data center capacity. All those elements are important, but the ability to operate them effectively depends on something less visible: the quality of the information connecting them.

AI may be the technology transforming the data center, but infrastructure intelligence is what allows operators to understand, plan and manage that transformation. The question is no longer simply, “How much more power, cooling or space do we need?” It’s also: “What do we have, what can it actually support, and what happens if we change it?” Without reliable answers, even the most advanced AI infrastructure can be constrained by the environment supporting it. 

With accurate, connected infrastructure intelligence, operators have the visibility and context needed to make faster, more informed decisions as AI continues to reshape the data center.

# # #

About the Author

Oliver Lindner has over 30 years of experience in IT and the management of IT infrastructures with a focus on data centers. He has worked for many years at FNT Software, a leading provider of software for managing digital infrastructure. In his current position as Director of Product Management, he is responsible for the strategic direction and continuous improvement of the software products for data centers. The aim is to support customers in the efficient and transparent design of their IT infrastructure.

Oliver Lindner attaches great importance to customer focus, innovation and quality. His expertise also includes the development and provision of Software as a Service (SaaS) solutions that offer customers maximum flexibility and efficiency. To this end, he works closely with his own team, partners and customers to create sustainable and innovative software solutions.

The post AI Is Only as Strong as the Infrastructure Intelligence Behind It appeared first on Data Center POST.

TL;DR

• Drawings Reflect Design Intent, Not Installed Reality: Equipment records extracted from engineering drawings show intended tag names and ratings, but they cannot confirm actual field swaps, serial numbers, or procurement updates.
• Separate Document Data from Field Verification: Every record entry should cite its source drawing number and revision letter, maintaining a clear distinction between drawing-derived data and field-confirmed status.
• Preserve Exact Tag Names and Explicitly Flag Uncertainty: Record tag names precisely as printed on drawings while logging operational designations separately, and explicitly mark ambiguous or unreadable values instead of making assumptions.

# # #

Before a data center is handed over, someone has to produce the equipment record that operations will live with: every switchboard, UPS module, generator, chiller, pump, CRAH unit and valve, with a tag, a location and a source. Most of that record starts in the drawings. One-line diagrams, mechanical piping schematics and cooling P&IDs name the equipment long before it is installed. Errors can enter the record when information is copied out of those drawings into a list.

A drawing can establish less than people assume. A one-line diagram shows that a breaker exists, what it feeds and what its tag is. It often shows a rating. It does not show the serial number, the firmware version, or whether the unit installed is the unit that was drawn. A cooling schematic shows a chilled water pump, its tag and the loop it serves. It does not show that the pump was swapped for a different model during procurement. A list built from drawings reflects what those documents record. It does not, by itself, verify what is installed. Treating it as a verified record of installed equipment is an easy mistake to make because the two can look identical in a spreadsheet.

The practical fix is to keep document-derived information and field verification distinct. Every row extracted from the documents carries the drawing number and the revision it was read from. A second step, done on the floor or from commissioning records, confirms each row against the installed equipment and adds what only the field can know. When the two disagree, the record should show the disagreement, with both values, until someone resolves it. A field marked “not shown on the drawing” is more useful to a facilities team than a plausible value nobody can trace.

Revisions are where document-derived lists drift. Power and cooling drawings change through design, construction and commissioning, and a list typed from revision B is quietly wrong by revision E. The tag of a renumbered panel changes on the drawing and not in the list. A generator added for a later phase appears on the one-line and never reaches the asset register. The habit that prevents this is simple and tedious: every row keeps the revision it came from, and each new drawing issue is compared with the last one, tag by tag, so that additions, deletions and renumbering are found as a set and not discovered one at a time after handover.

Tags deserve their own care. The same piece of equipment is often named three ways: the designer’s tag on the drawing, the manufacturer’s designation on the nameplate, and the name the operations team wants in the maintenance system. Problems start when a list silently translates one into another. Keep the tag exactly as the drawing prints it, hyphens and all, and add the other names beside it as separate fields. Anyone who later has to trace a work order back to a drawing will be able to.

Uncertainty should be written down, not smoothed over. Scanned drawings have characters that cannot be read with confidence. A tag may be partly hidden by a revision cloud, or a rating may be legible on one sheet and different on another. The person preparing the list will be tempted to pick the likely answer and move on. A better record marks the row as unconfirmed, says why, and points at the sheet. The commissioning team can then prioritize flagged rows while carrying out the required verification of the wider record.

A short checklist covers most of it. Every row names its source drawing and revision. Tags are recorded as printed. Document-derived information and field verification status are distinguishable. Unreadable or conflicting values are flagged. Each new drawing issue is compared against the list before the list is reissued.

The principle is straightforward. Every line in an equipment record should show where it came from and whether it has been verified. Data centers are built quickly and handed over on a date, and that record is one of the few things from the project that the operations team will still be using ten years later.

# # #

About the Author

Muaawia Jannat is the founder and CEO of Tagsight Technologies Inc., a Calgary-based company developing engineering software for instrumentation and controls teams. His work focuses on turning engineering drawings into structured equipment and instrument records that teams can review, trace to their source and use in downstream workflows.

The post An Equipment List Is Only as Useful as Its Source appeared first on Data Center POST.

TL;DR

• Power Readiness Scheduling: Grid interconnection wait times reaching five to ten years in many regions require maintaining a dedicated power-readiness schedule that tracks utility upgrades, energization permissions, and phased deployment dependencies.
• Rack Level Cooling and Water Strategy: Facility engineering must look beyond hall averages to accommodate high rack densities with direct liquid cooling, traceable heat paths, and site-specific water management plans.
• Shared Acceptance Decisions: AI platform, facilities, and energy procurement leaders must jointly approve site readiness under real-world operating conditions, explicitly pricing the financial risk of delayed or restricted operations.

# # #

Picture an AI project with everything lined up: accelerators secured, lease signed, installation window booked. Then the utility’s service date slips, or the cooling design cannot support the specified racks through a hot afternoon with a pump unavailable. The hardware exists, but the computing capacity the business expected does not. Buying the machines was never the same as buying the ability to run them.

That distinction should change how AI infrastructure is planned. Chip availability still matters, but a site can become the binding constraint long before its floor space is full. Power delivery, heat removal and workload commitments need to be settled together, while there is still time to change the location or deployment plan.

Even without pinning every data center kilowatt-hour on AI, the scale is substantial. The IEA’s April 2026 outlook estimates global data center electricity consumption at about 485 terawatt-hours in 2025 and projects roughly 950 TWh in 2030, nearly double. Those totals cover data centers broadly; the report separately projects electricity use by AI-focused facilities to triple over that period. The 2030 figures are projections, not measured demand.

Start with a deliverable service date

A power discussion needs a location, a date and operating conditions, not just a megawatt figure. In its 2026 report, the IEA identifies grid-connection waits reaching five to ten years in many jurisdictions. That is a warning about constrained markets, not a universal lead time to paste into every project schedule.

Queue statistics also require care. Berkeley Lab’s 2026 Queued Up analysis reports a median exceeding five years from interconnection request to commercial operation for projects completed in 2025, in regions with available data. It tracks generation and storage projects, not data center load connections. Treating that number as the average wait for an AI campus would misstate the evidence.

The rules for connecting large loads are also in flux. In June 2026, FERC directed six regional grid operators to justify or revise their tariffs for large loads, addressing studies, cost allocation, co-location and flexible service. That action was not an energization approval for any individual campus. Developers still need to establish what their particular service agreement permits and what infrastructure must be completed first.

My recommendation is to maintain a power-readiness schedule alongside the construction schedule. It should name the dependencies for each deployment phase: studies, utility upgrades, equipment delivery, protection testing and permission to energize. Ask which dates are commitments, which are estimates and who carries the cost of slippage. A smaller first phase with power you can count on may beat a larger phase whose power date is still conditional.

Design around the demanding rack rather than the average hall

Once electricity reaches the building, the challenge becomes more local. Two hypothetical halls could have identical IT megawatts while placing very different demands on individual rows, electrical branches and cooling loops. Averaging across the hall would hide that difference. Require the proposed rack layout and expected workload profile before accepting a cooling-capacity claim.

Direct liquid cooling moves heat into fluid close to where it is generated, rather than first moving it through room air. The Department of Energy’s 2024 design guide describes its ability to serve higher heat densities. But selecting liquid-cooled servers is not the same as specifying the facility’s entire heat-removal system.

Temperature compatibility deserves particular attention. ASHRAE’s liquid-cooling guidance explains that increasing component power can require lower supply-water temperatures, potentially reducing opportunities to cool without mechanical refrigeration. A loop suitable for one hardware generation is therefore not automatically suitable for the next. Validate the actual equipment envelope rather than assuming that “liquid-ready” settles the question.

For design reviews, I would ask for a traceable heat path from the chip to the outside environment. Which components remain air-cooled, what temperatures and flow rates are required, and how does capacity change during maintenance? Include coolant distribution units, pumps, piping, controls and leak response in commissioning scenarios. A cooling system should earn its operating limits through testing, not through a label on the sales specification.

Keep efficiency claims within their measurement boundaries, too. DOE defines power usage effectiveness, or PUE, as annual facility energy divided by annual IT energy; it measures infrastructure overhead, not useful computing output. A favorable annual PUE does not establish peak electrical or thermal capacity. For an AI deployment, I would review it alongside completed work, service quality, total energy and peak demand.

Water belongs in the siting decision

Water accounting has two distinct operational boundaries. Berkeley Lab’s 2024 U.S. data center report estimates 2023 direct water consumption at about 66 billion liters and the indirect footprint from electricity at nearly 800 billion liters. These are sector-wide estimates, not AI-only measurements. The indirect estimate uses regional electricity mixes and excludes facility-specific power-purchase agreements and behind-the-meter generation.

Liquid cooling does not, by itself, tell you how much water a facility consumes. In December 2024, Microsoft described a new design using recirculating water without evaporation for cooling, while acknowledging increased energy demand when mechanical cooling replaces evaporation. That was a design announcement, not evidence that its whole fleet had eliminated cooling-water use. Nor does zero evaporation at the site erase water consumption associated with electricity generation.

The local source matters as much as the annual total. Google’s published water-risk framework considers community demand, available supply, delivery infrastructure and changing climate conditions. It describes choosing air cooling for Mesa, Arizona, after identifying elevated risk to the proposed water source. That is a site-specific decision, not a universal endorsement of one cooling architecture.

For a new project, request a water plan before making the final siting commitment. Separate withdrawal from consumption, identify potable and reclaimed supplies, and test the plan against seasonal restrictions and extreme heat. Ask what happens to electrical demand if water availability falls. A design that saves electricity by using water needs a credible operating plan for the days when that water is hardest to obtain.

Buy flexibility only when the workload can deliver it

There is public evidence that some computing demand can move. In August 2025, Google announced utility agreements targeting machine-learning workloads for demand response and reported reducing associated demand during three grid events in 2024. It also described shifting non-urgent tasks. These are specific operator disclosures, not proof that every AI workload can be interrupted economically.

I would qualify flexibility workload by workload. A batch job with deadline slack might be deferrable; an interactive service with a tight response-time commitment may not be. For training, examine checkpoint overhead, restart time and the cost of leaving accelerators idle. Moving work elsewhere also requires spare capacity, sufficient connectivity and permission to move the relevant data.

The useful deliverable is a tested operating specification: how many megawatts can be reduced, with how much notice, for how long, and at what recovery cost? Measure the reduction at the facility boundary rather than assuming that an application-level pause produces an equal grid benefit. Test the recovery period as well; deferred work still needs somewhere and some time to run. Do not promise flexibility to an energy supplier before the service owners have accepted those constraints.

Make site readiness a shared acceptance decision

Before approving an accelerator purchase, I would bring the AI platform lead, facilities engineer and energy-procurement lead to the same review. Their shared acceptance record should answer three questions:

• What capacity can be delivered at commissioning, and which electrical or thermal conditions reduce it?
• What evidence supports the next expansion date, and what happens if a dependency slips?
• Who owns each unresolved constraint, including decisions that cut across facilities, software and energy procurement?

Compare candidate locations against the same workload, including the earliest usable power date, supported rack envelope, water restrictions, curtailment obligations and network constraints. Then price a delayed deployment and a restricted operating year, rather than comparing only steady-state electricity costs. If the preferred site wins only when every upgrade arrives on its earliest estimated date, examine the downside before discarding the alternative. This comparison also gives the business a concrete choice: wait, deploy less, change the workload or move.

The central planning question is straightforward: what work can this site reliably deliver under the conditions in which it will actually operate? A hardware count cannot answer it. Neither can an annual energy contract, a good PUE number or a promise of future cooling upgrades. AI capacity becomes real when those pieces work together.

The views expressed here are my own. This article does not represent the views, policies, practices, or technology stack of my employer, LinkedIn. All examples are based on public sources or explicitly hypothetical scenarios. 

# # #

About the Author

Silu Panda is a staff software engineer at LinkedIn, focused on AI infrastructure and model serving. His interests include reliable inference, distributed systems, and open source. He writes about evaluating, deploying, and operating AI applications. 

The post AI Infrastructure Is Running Into Power Grid and Cooling Limits appeared first on Data Center POST.

TL;DR

• Expanded network reach: Enables service providers to deliver private Layer 2 Ethernet connectivity beyond their existing fiber footprints without additional fiber construction.
• Automated connectivity: Pairs with the MaiaEdge PBC 2000 to establish encrypted Ethernet connections over available fiber, dedicated internet access, satellite and other IP transport.
• New service opportunities: Supports cloud on-ramps, distributed AI access and additional services through existing customer connections.
• Faster customer onboarding: Reduces manual network configuration for service providers, data center operators and AI infrastructure companies.

###

MaiaEdge has introduced its PBC 100 Series, compact customer premises equipment (CPE) designed to help fiber operators and service providers extend private Ethernet connectivity beyond their existing fiber networks. The new solution enables operators to reach off-net customers without additional fiber construction or manually negotiated network-to-network interfaces (NNIs).

Part of MaiaEdge’s Path Border Controller (PBC) product family, the PBC 100 Series works alongside the company’s PBC 2000, deployed at an operator’s point of presence (PoP), Metro Ethernet edge or data center. Together, the devices enable automated, encrypted Layer 2 connectivity across available fiber, dedicated internet access, satellite and other IP transport.

Built on MaiaEdge’s SD-Ethernet architecture, the PBC 100 Series extends automated path provisioning directly to customer premises. A centralized interface simplifies routing complexity and provides operators with visibility across the connection, reducing the time typically required to establish services from weeks or months.

“For decades, where your fiber ended, your service opportunity ended with it,” said Abilash Menon, CEO and Co-Founder of MaiaEdge. “The PBC 100 Series breaks that link. Providers can reach off-net customers without building fiber or negotiating a new NNI, then turn that connection into private connectivity, cloud access, AI infrastructure, and more. We’re giving operators a way to expand their addressable market without expanding their physical network.”

The PBC 100 Series also enables operators to introduce additional services through the same customer connection, including cloud on-ramps and distributed AI access. For data center and AI infrastructure operators, the technology can accelerate tenant onboarding by reducing manual network configuration while creating opportunities to deliver services through their own networks or MaiaEdge-enabled partners.

The launch expands MaiaEdge’s Path Border Controller portfolio, allowing operators to extend private Ethernet services directly to customer locations while maintaining their existing physical network infrastructure. The first device in the series, the PBC 124-C, is available now.

To learn more about MaiaEdge and the PBC 100 Series, visit maiaedge.io

The post MaiaEdge Introduces PBC 100 Series to Extend Private Ethernet Beyond Fiber Networks appeared first on Data Center POST.

TL;DR

• Growing Facility Complexity: Rapidly rising rack densities and liquid cooling demands are outpacing traditional construction and labor availability.
• Cascading Delay Costs: A missed milestone affects financing costs, contractual penalties, and future capacity commitments.
• Discipline Around Time: Achieving completion certainty requires early risk visibility and treating schedule management with the same discipline as financial capital.

# # #

Data center development has always been complex, but what has changed is the scale.

During a recent Yotta panel, Dr. Atif Ansar, Executive Chairman and Co-Founder of Foresight, alongside industry leaders, discussed how projects are becoming larger, denser and more technically demanding while many of the resources required to deliver them have not expanded at the same pace. Skilled labor remains constrained, power is increasingly difficult to secure, equipment lead times remain significant and infrastructure requirements continue to evolve.

As a result, the consequences of a missed milestone are growing. A delay rarely stays isolated to one part of a project. It can ripple across contractors, suppliers, customers, financing and future capacity.

Complexity Is Increasing Faster Than Delivery Capacity

Not long ago, a 100 or 150 MW data center represented significant scale. Today, the industry is developing campuses measured in hundreds of megawatts and even gigawatts. Rack densities are also climbing rapidly, while direct liquid cooling and other new infrastructure requirements are reshaping facility design.

Meanwhile, the labor pool has not grown at the same rate as construction demand. Supply chains and power availability add further pressure.

This leaves less room for delays, design changes or slow decision-making.

Managing the Entire Delivery Chain

Data center schedules extend well beyond construction. Land, power, design, procurement, manufacturing, logistics, commissioning and customer readiness all contribute to the final ready-for-service date.

Panelists discussed how vertical integration, modular construction and offsite manufacturing can help reduce field labor and provide greater control over schedules. But improving delivery also requires stronger coordination between owners, contractors, customers and suppliers.

For owners, that does not mean managing every detail of construction. It means maintaining visibility into the milestones that matter and understanding what needs to happen today to protect a delivery date months into the future.

Time Has a Direct Financial Impact

The consequences of delay extend beyond construction costs. Facilities that come online late can face lost revenue, contractual penalties, increased financing costs and delayed customer capacity.

During the panel, Dr. Atif Ansar illustrated the potential impact using a 100 MW data center, estimating approximately $20 million in monthly revenue exposure before additional delay-related costs are considered.

That makes early visibility critical. The sooner teams identify schedule risk, the more opportunity they have to adjust resources, resequence work or coordinate with customers and suppliers.

Delivery Certainty Is Becoming a Competitive Advantage

The industry cannot eliminate uncertainty. Technology will continue to evolve, labor and power constraints will remain, and supply chains will face disruption.

The opportunity is to manage those challenges earlier and more effectively.

Successful delivery will increasingly depend on shared visibility, faster decision-making and stronger alignment across the project ecosystem. At this scale, time deserves the same discipline and attention organizations already give to capital.

The post Why Data Center Scale Is Raising the Stakes for Project Delivery appeared first on Data Center POST.

TL;DR

• More than 1,500 new data centers are in various stages of development across the U.S., but developers are encountering growing opposition related to electricity costs, water and energy use, environmental impacts and artificial intelligence.
• States and municipalities are responding with new requirements, moratoriums and bans, forcing developers to reconsider when and how they engage local communities.
• Industry leaders say transparency, early outreach and tangible community benefits are becoming essential to advancing responsible data center projects.

###

More than 1,500 new data centers are in various stages of development across the United States, but growing community opposition is changing how projects move forward. According to Inside Towers Intelligence’s 2026 Volume 3 Digital Infrastructure Industry: Market Analysis Report, concerns about electricity costs, energy and water use, environmental impacts, and artificial intelligence are driving greater public scrutiny of proposed developments.

Public resistance is widespread. A Gallup poll conducted in March 2026 found that seven in ten Americans opposed data center construction in their local communities including more than half who strongly opposed it. Environmental concerns are particularly significant with 55% of U.S. adults saying they worry a great deal about data centers’ environmental impact and another 23% expressing a fair amount of concern.

Government officials are responding with a growing patchwork of regulations, moratoriums, and bans. Several states have considered restrictions on new developments, while numerous municipalities have enacted or explored local pauses. These actions are already affecting project timelines and encouraging developers to reevaluate their site-selection and community-engagement strategies.

As opposition increases, early community outreach is becoming a prerequisite for development. Instead of securing land, power, fiber, and water before speaking with residents, developers are increasingly engaging local leaders and communities earlier in the planning process. These conversations allow companies to explain their projects, address local concerns, and identify benefits that can be delivered to the surrounding area.

Several developers are pairing transparency with tangible community investments. Examples highlighted in the report include support for schools, roads, workforce-development programs, and environmental initiatives. Duos Edge AI also hosts open houses and guided facility tours that allow residents, educators, officials, and businesses to see its modular edge data centers and ask questions directly about their size, cooling systems, and electricity requirements.

Clearer communication will remain central as the industry responds to concerns about power, water, jobs, and the environment. Ilissa Miller, CEO of iMiller Public Relations, recommends approaching community engagement more like an election campaign than a traditional project-marketing effort. She also advocates for engaging communities before filing conditional-use applications and providing clear information about what developers can offer at each stage of the process.

The report indicates that responsible development will increasingly depend on transparency, thoughtful planning, and measurable community value. Operators that engage early, correct misinformation, and work constructively with residents and local officials will be better positioned to move projects forward. Explore the Inside Towers Intelligence report for additional insights into the community and regulatory pressures shaping data center development.

The post Early Community Engagement Becomes Critical as Data Center Opposition Grows appeared first on Data Center POST.

TL;DR

• Unified Infrastructure Platform: MOD Mission Critical’s PaaSPort consolidates fragmented vendor relationships for colocation, bare metal, connectivity, and cloud access into a single global platform.
• Comprehensive Global Services: The platform offers on-demand hardware, IP transit scaling from GigE to 10G across six continents, virtual hosting, IPv4 space, and network monitoring through an open provider ecosystem.
• Addressing Edge Expansion: Unified edge solutions help enterprises manage growing complexity as the global edge computing market expands from $84 billion to over $1.5 trillion by 2035.

# # #

Written by iMiller Public Relations on behalf of MOD Mission Critical.

Traditional digital infrastructure deployment often forces organizations to stitch together a patchwork of separate vendors for colocation, connectivity, bare metal, and cloud access. Managing these isolated services introduces vendor sprawl, multiple contracts, tracking separate SLAs, and coordinating complex escalation paths during outages. To solve these operational challenges, MOD Mission Critical developed PaaSPort, a unified platform designed to bring global infrastructure capabilities together under a single ecosystem. Building on capabilities developed since 2011, MOD enables businesses to explore, compare, and price global infrastructure options through a single partner.

The PaaSPort platform provides access to a comprehensive suite of infrastructure services across six continents. Organizations can deploy on-demand physical bare metal hardware, vendor-agnostic colocation procurement, and IP transit scaling from GigE to 10G. Additionally, the platform covers virtual hosting, dedicated hosting, IPv4 services, managed proxy, and network monitoring. By offering an open ecosystem for both customers and infrastructure suppliers, the platform streamlines multi-market sourcing into a single procurement framework.

Simplifying distributed infrastructure management is increasingly vital as enterprise edge deployments expand worldwide. Industry projections indicate that over 60 percent of enterprises globally will adopt unified edge frameworks by 2027 to control operational complexity. Furthermore, the global edge computing market is projected to surge from approximately $84 billion today to over $1.5 trillion by 2035. As infrastructure requirements grow across global markets, unified sourcing platforms help organizations scale their footprints without accumulating administrative friction.

To read the original article, please click here.

The post PaaSPort: Bringing MOD’s Global Infrastructure Services Into One Platform appeared first on Data Center POST.

TL;DR

• Grid Constraints Suppress Pipeline: Announced US data center power demands exceed peak grid capacity, but interconnection delays will prevent most projects from being built
• Behind-the-Meter Power Shift: With 40 percent of AI data centers facing power limits by 2027, operators are turning to on-site generation to bypass long queue times
• Blind Efficiency Verification: Skepticism over unverified vendor energy claims requires predictions against withheld operational data to prove performance

# # #

Written by iMiller Public Relations on Behalf of Physical Superintelligence (PSI).

Announced data center projects in the United States total roughly 780 gigawatts, surpassing the country’s all-time peak load of 759 gigawatts set in July 2025. Although market demand for compute power remains high, PSI’s Scott Charter sheds light on the factors keeping a significant portion of this planned pipeline from materializing; chief among them, critical power availability and interconnection bottlenecks. Across nearly every market outside China, the central obstacle to expanding compute capacity lies in grid transmission backlogs rather than a willingness to construct power generation. 

To avoid multi-year interconnection queues, expansion in terrestrial compute through 2029 will increasingly adopt behind-the-meter, on-site generation. This transition aligns with projections showing that 40 percent of existing AI data centers will be operationally constrained by power by 2027, with global compute infrastructure likely hitting a definitive power wall within five to seven years. 

Simultaneously, vendor energy efficiency claims face widespread skepticism due to a market for lemons dynamic, as unverified percentage metrics prompt buyers to discount assertions to zero. Overcoming this skepticism requires rigorous verification, such as submitting predictions against withheld data from specific halls or operational periods before those datasets are revealed

To read the original article, please click here.

The post America Has Announced More Data Centers Than It Has Grid appeared first on Data Center POST.

TL;DR

• Customer continuity: Service at Columbus continues without interruption under a revised five-year agreement.
• Expansion focus: Redirects capital from GPU ownership toward new modular AI colocation sites.
• Continued service: Keeps the Columbus customer supported without interruption under an enhanced five-year agreement.
• Financial flexibility: Eliminates approximately $98.1 million in prospective GPU equipment financing, preserving capacity for site development.

# # #

Duos Technologies Group, Inc. (“Duos” or the “Company”) (Nasdaq: DUOT) has completed the sale of its GPU-as-a-Service entity, Duos Edge AI – GPUaaS, LLC, to Axe Compute, Inc. (Nasdaq: AGPU). The transaction closed on September 30, 2026, advancing Duos’ shift toward a focused AI colocation platform.

Duos will continue to own and operate its Columbus, Georgia colocation facility, providing power, cooling, security, and managed infrastructure services. Rather than owning and financing the GPUs, the Company will lease the underlying compute capacity from Axe Compute. Its existing Columbus customer will continue receiving service without interruption under a revised five-year agreement expected to increase the revenue Duos recognizes over the contract’s life.

Under the agreement, Axe Compute acquired all outstanding equity interests in the entity holding Duos’ GPU compute cluster and associated equipment supply and financing positions. Duos will receive $42.9 million through evenly scheduled monthly payments over the next 60 months, while approximately $98.1 million in prospective equipment financing obligations have been removed.

“Selling the GPU fleet sharpens Duos into a pure-play edge data center operator,” said Chris DeAlmeida, Duos CFO. “We keep the customer and the economics, we shed roughly $98.1 million of prospective equipment debt, and we free the balance sheet to fund sites instead of servers. Our outlook for 2026 is unchanged: full year revenue above $50 million and positive adjusted EBITDA.”

The transaction allows Duos to direct capital toward developing and operating 10 to 30 MW modular colocation sites across the United States. Its broader strategy centers on owning the sites, power capacity, and infrastructure that support AI workloads, while customers bring and operate their compute equipment.

The sale reflects Duos’ continued focus on building long-lived colocation assets and reinvesting capital into new sites, campuses, and customer deployments.

To learn more about Duos Technologies Group, visit duostechnologies.com.

The post Duos Technologies Group, Inc. Sells GPU-as-a-Service Entity to Axe Compute appeared first on Data Center POST.

TL;DR

• Introducing SD-Ethernet: Software-Defined Ethernet transforms Ethernet into an automated, cross-domain service that is encrypted by default and programmable via APIs.
• MaiaEdge Architecture: By pairing Path Border Controllers with a cloud-based Path Computation Engine, MaiaEdge slashes service activation times from 60 to 90 days to seconds while offering real-time telemetry across segments.
• Built for AI Demands: SD-Ethernet enables enterprises to instantly extend their private network directly to remote GPU clusters and clouds, converting static network pipes into a dynamic platform for AI workloads.

# # #

Written by iMiller Public Relations on behalf of MaiaEdge.

Over five decades, the networking industry introduced a succession of protocols and architectures, from ATM and MPLS to SD-WAN and SASE, to address complex connectivity, routing, and security challenges. Despite these evolving abstractions and overlay networks, Ethernet remained the universal foundation at the port level, natively used by servers, switches, clouds, and GPU clusters. However, traditional Ethernet services suffered from severe operational bottlenecks, requiring manual cross-connect contracts, spreadsheets, and lengthy negotiations that delayed multi-domain provisioning for 60 to 90 days.

To eliminate these legacy friction points, Software-Defined Ethernet (SD-Ethernet) introduces a programmable, automated layer that extends secure Ethernet handoffs across independent operators, fiber networks, and cloud providers. Utilizing MaiaEdge’s Path Border Controllers at network boundaries and a cloud-based Path Computation Engine, the architecture dynamically computes cross-domain routes, encrypts traffic end-to-end, and measures performance along every segment. This approach slashes connection activation times from months down to seconds while delivering complete path visibility, allowing organizations to pinpoint the exact provider domain responsible for packet loss.

This paradigm shift is heavily driven by the demands of AI infrastructure, where distributed training models, inference nodes, and sovereign datasets require fast, private, and flexible interconnections. Rather than forcing enterprises to move into entirely new networks or manage fragmented security stacks, SD-Ethernet allows existing private networks to seamlessly extend wherever GPU resources reside. Consequently, networking transitions from static, slow-to-provision pipes into an agile platform built for high-performance AI data movement.

To read the original article, please click here.

The post Why Ethernet Never Left: A 50-Year Networking Mystery appeared first on Data Center POST.

TL;DR

• Broadband strategies are expanding beyond network construction to address adoption, affordability, digital skills, and meaningful use.
• Community-led partnerships are helping align infrastructure investment with local priorities, including Tribal, rural, and historically underserved markets.
• Public and private funding must be structured around long-term operations, accountability, and sustainability, not simply initial deployment.
• Multifamily connectivity decisions increasingly depend on resident expectations, practical technology planning, policy requirements, and measurable performance.

# # #

Broadband Communities Summit 2026 placed a clear emphasis on what happens after infrastructure reaches a community. Held August 25-27 in Houston, Texas, the event examined how broadband providers, municipalities, Tribal organizations, property owners, investors, and nonprofit leaders can turn network availability into reliable, affordable, and useful connectivity. Across the Communities, Finance & Frameworks, Infrastructure and Multifamily tracks, the agenda reflected an industry that is balancing ambitious deployment goals with the practical work of adoption, operations, and long-term value.

The summit also highlighted that broadband challenges vary widely by market. Rural communities may need flexible technology and funding models, multifamily owners must weigh resident experience against infrastructure readiness and regulatory demands, and local leaders must build trust among providers, utilities and community organizations. The strongest strategies discussed in Houston treated connectivity as an ecosystem, one that depends on capital, technology, policy, and community participation working together.

Community Needs Must Shape Broadband Strategy

Access remained essential, but summit programming repeatedly looked beyond coverage maps. Sessions explored affordability, device availability, digital literacy, and the role of local organizations in helping residents use broadband for education, healthcare, employment, and economic opportunity. That broader view positions adoption and measurable community benefit as core parts of network planning rather than secondary considerations after construction.

In the keynote “Digital Justice: Reimagining Connectivity for All,” Francella Ochillo, Adjunct Professor and Consultant at Johns Hopkins University, addressed the evolving broadband landscape and the strategic approaches required to overcome persistent industry hurdles. The event’s community-focused sessions reinforced the same principle: successful broadband initiatives begin with accurate local data and sustained engagement, then connect infrastructure decisions to the needs of the people the network is intended to serve.

Local and Tribal Partnerships Create New Infrastructure Pathways

Partnership models were especially prominent in sessions focused on rural, regional and Tribal connectivity. Rather than relying on a single organization to finance, build and operate every component, communities are assembling coalitions that combine local knowledge, public resources, private-sector expertise and shared infrastructure. These arrangements can expand access while giving communities a stronger voice in how networks are designed and governed.

Joel Ogren, Chief Executive Officer of Assured Communications, joined Tyson Johnston of Toptana Technologies for “Building Connectivity from the Coast,” a presentation on a Tribal partnership advancing subsea and backhaul infrastructure. The session illustrated how major connectivity projects can link technical deployment with regional priorities. It also highlighted the value of partnership when projects must coordinate difficult geography, long-haul infrastructure and the economic goals of the communities involved.

Financing Must Support the Full Network Lifecycle

The summit’s finance programming made clear that securing capital is only one part of a viable broadband plan. Workshops and panels addressed private equity expectations, commercial debt, public-private partnerships, grant compliance, and approaches to layering public and private investment. Across these topics, the central question was whether a project’s financial structure could support construction, ongoing maintenance, service quality, and future growth.

Kyle Zimbelman, Chief Operating Officer of Golden State Connect Authority, participated in “Public-Private Partnerships in Practice,” which examined how partners can align state, local, and provider priorities while balancing accountability, flexibility, and speed. The discussion framework emphasized careful partner selection and sustainable operating models which are important considerations as communities move from funding awards and project announcements to execution and long-term stewardship.

Community-Led Models Can Turn Ambition Into Adoption

Several sessions offered practical examples of communities pursuing new approaches where conventional deployment models had fallen short. These case studies showed that local leadership, coordinated outreach, and a clearly defined public mission can help projects navigate political, financial, and operational barriers. They also demonstrated why success should be evaluated through adoption and community outcomes, not only through miles of fiber or premises passed.

Joshua Edmonds, CEO of DigitalC, presented “From Impossible to Improbable to Impactful: Exploring the Rise of the Cleveland Model.” His session traced the development of an approach designed to deliver quality and affordability in one of the nation’s least-connected cities. The Cleveland experience provided a useful example of how a community-centered operating model can evolve when technology, political commitment, and execution capacity are built around a long-term effort to close the digital divide.

Multifamily Connectivity Requires Practical, Data-Informed Decisions

The Multifamily track addressed another rapidly changing part of the broadband market. Property owners and service providers examined Wi-Fi standards, resident expectations, bulk broadband and video adoption, implementation challenges, fee transparency, and the operational effects of new policies. The sessions reflected a growing need to translate technology and regulation into decisions that make sense at the property level.

Elizabeth Parks, President and CMO of Parks Associates, contributed to the Multifamily Town Hall and a data-focused session on resident satisfaction and adoption trends. That market perspective supported one of the track’s clearest messages: owners need realistic performance benchmarks and credible adoption data when planning conversions, evaluating infrastructure upgrades, and assessing the resident experience. Technology choices create lasting value when they match building conditions, operational capacity and demonstrated customer needs.

Building Broadband That Lasts

Broadband Communities Summit 2026 reflected an industry entering a more demanding phase. Funding programs and new technologies continue to create opportunities, but lasting results depend on whether communities and providers can coordinate deployment, earn trust, support adoption, and operate networks effectively over time. The most durable plans will connect financial discipline and technical performance with clearly defined community outcomes.

As broadband stakeholders move forward, the Houston discussions offer a useful standard for evaluating progress: infrastructure matters, but its value is ultimately measured by the people, properties, and local economies it serves. To learn more about Broadband Communities Summit and future events, visit the official Broadband Communities Summit website.

The post Broadband Communities Summit 2026 Highlights the Shift From Network Access to Lasting Community Impact appeared first on Data Center POST.

TL;DR

• Traditional backup is no longer sufficient for modern data centers due to rapid data growth, complex hybrid IT environments, slow recovery times, and sophisticated ransomware threats.
• Resilient protection enhances core frameworks like the 3-2-1 backup strategy by incorporating immutable storage, air-gapped backups, encryption, and automated recovery testing.
• Purpose-built appliances simplify operations by combining storage hardware, software integration, and hardened security into a single platform for faster deployment and reliable recovery.

# # #

Data centers are the foundation of modern digital businesses. From financial services and healthcare to SaaS platforms and manufacturing, organizations depend on always-available infrastructure to run critical workloads.

However, as data volumes continue to grow and cyber threats become more advanced, traditional backup approaches are no longer enough. Organizations need backup strategies that provide fast recovery, ransomware protection, operational simplicity, and long-term data availability.

A modern data center backup strategy must address not only hardware failures but also cyberattacks, accidental deletion, insider threats, and increasingly complex hybrid IT environments.

The Growing Challenges Facing Data Center Backup

Data protection has become more complicated because organizations now manage data across multiple environments, including:

• On-premises data centers
• Private clouds
• Public cloud platforms
• Virtualized environments
• Remote offices and edge locations

At the same time, ransomware attacks have changed the requirements for backup infrastructure. Simply having copies of data is no longer enough. Organizations need protected backups that attackers cannot easily modify or encrypt.

This shift has moved backup from being a simple storage function to becoming a critical component of business continuity planning.

Why Traditional Backup Methods Are No Longer Enough

Legacy backup systems were designed primarily for recovery after hardware failures or accidental data loss. While they remain useful, they often lack the capabilities required for modern threats.

Common limitations include:

Slow Recovery Times

Large enterprises may need to restore terabytes or even petabytes of data after an incident. Traditional backup systems can result in long recovery windows, increasing downtime costs.

Limited Protection Against Ransomware

If backup copies remain connected and writable, attackers may target them during an attack. Organizations increasingly require immutable and isolated backup copies.

Complex Management

Many enterprises operate multiple backup tools across different environments. This creates management overhead and increases the possibility of configuration errors.

Building a Strong Data Backup Strategy

A reliable backup architecture should follow established data protection principles. One widely adopted approach is the 3-2-1 backup strategy, which recommends maintaining three copies of data, storing them on two different types of media, with one copy kept offsite.

This approach helps organizations reduce risks caused by hardware failures, disasters, and cyber incidents.

However, modern enterprises often extend this model by adding additional protection layers such as:

• Immutable storage
• Air-gapped backups
• Encryption
• Automated recovery testing
• Continuous monitoring

The Role of Purpose-Built Backup Appliances

As data environments become more complex, many organizations are moving toward purpose-built backup infrastructure instead of relying only on general-purpose storage systems.

A backup appliance combines storage hardware, backup software integration, and data protection features into a single platform.

For data centers, this approach provides several advantages:

• Faster deployment
• Simplified management
• Optimized backup performance
• Better integration with enterprise applications
• Improved recovery reliability

Purpose-built solutions can also support advanced protection features such as immutable backups and ransomware-resistant storage architectures.

Immutable Backup for Modern Data Protection

One of the biggest changes in enterprise backup is the adoption of immutable storage.

Immutable backups prevent stored backup data from being altered or deleted during a defined retention period. This capability is especially important for organizations facing ransomware threats.

For environments using Veeam, organizations often deploy a Veeam immutable backup appliance to create protected backup repositories that support secure recovery after cyber incidents.

Combining backup software with hardened storage infrastructure creates a stronger defense against modern threats.

Disaster Recovery Requires More Than Backup Copies

Backup and disaster recovery are closely connected but serve different purposes.

Backup focuses on protecting copies of data, while disaster recovery focuses on restoring business operations after an outage or failure.

A complete disaster recovery strategy should consider:

• Recovery Point Objectives (RPO)
• Recovery Time Objectives (RTO)
• Application dependencies
• Infrastructure availability
• Geographic risks

Many organizations now use a disaster recovery appliance to simplify recovery processes and reduce downtime during unexpected events.

Cloud and Hybrid Recovery Models

Cloud infrastructure has introduced new possibilities for disaster recovery. Organizations can now combine local backup infrastructure with cloud-based recovery options.

A hybrid approach allows businesses to keep frequently accessed recovery data locally while using cloud resources for additional resilience.

However, cloud recovery strategies should still include careful planning around:

• Data transfer costs
• Recovery speed
• Compliance requirements
• Security controls

Cloud-based disaster recovery works best when integrated into a broader business continuity strategy.

Choosing the Right Backup Architecture for Your Data Center

Every organization has different requirements based on workload size, compliance needs, recovery objectives, and infrastructure design.

Before selecting a backup solution, IT teams should evaluate:

Data Growth

Rapidly increasing datasets require scalable backup infrastructure.

Recovery Requirements

Critical workloads may require near-instant recovery capabilities.

Security Requirements

Organizations should prioritize solutions supporting encryption, immutability, and access controls.

Operational Complexity

Simplified management helps reduce administrative overhead and improves reliability.

The Future of Data Center Backup

The future of backup will focus on automation, cybersecurity, and intelligent recovery.

Organizations are moving toward backup platforms that can:

• Detect unusual activity
• Automate recovery workflows
• Protect against ransomware
• Support hybrid environments
• Scale with growing data demands

For modern data centers, backup is no longer just a secondary storage function. It has become a core component of infrastructure resilience.

Conclusion

As organizations continue to depend on digital infrastructure, data protection must evolve beyond traditional backup methods. A modern approach combines proven strategies like the 3-2-1 backup strategy, immutable storage, disaster recovery planning, and purpose-built backup infrastructure.

By investing in resilient backup architectures, data centers can improve availability, reduce downtime risks, and maintain business continuity even during unexpected disruptions.

# # #

About the Author

George Williams, Senior Technical & Product Marketing Specialist, StoneFly Inc.

George Williams has over 13 years of experience in the data storage, backup and disaster recovery, air-gapping, immutability and archiving markets. A true geek with love for ease and simplicity in data storage, George has been working for StoneFly Inc. for over a decade. Ever since StoneFly started shipping products since 2006, George has been working to ensure that technical information is relayed in a simple and effective way to customers and targeted audience. George helps curate content and works with numerous publishers and technology blogs to spread awareness and knowledge of data storage technology.

The post Why Modern Data Centers Need a Resilient Backup Strategy Beyond Traditional Protection appeared first on Data Center POST.

TL;DR

• Equal Speeds Matter: Symmetrical internet delivers matching upload and download speeds, ensuring data sent out travels as fast as incoming data.
• Eliminates Bottlenecks: Strong upload capacity prevents video call freezing, slow file sharing, and background cloud backup delays when multiple family members are online.
• The Empire Advantage: Empire Fiber Internet provides equal upload and download speeds across connected household devices, and residents can check local address availability to upgrade.

# # #

Written by iMiller Public Relations on behalf of Empire Fiber Internet.

While traditional home internet plans prioritize download speeds for loading webpages and streaming shows, upload speed plays an equally critical role in household connectivity. Upload capacity governs everything sent from a home out to the internet, including participating in video calls, submitting school assignments, and backing up photos to the cloud. Symmetrical internet connections provide equal maximum speeds for both uploading and downloading data, delivering balanced bandwidth that fiber internet infrastructure inherently supports.

On busy school nights when multiple family members are online simultaneously, insufficient upload speeds create noticeable bottlenecks. Video calls can freeze or drop audio, group project uploads take longer, and automated cloud backups stall background productivity across smart household devices. Having equal upload and download capacity prevents high-bandwidth activities from interfering with one another, ensuring smooth performance for learning, working, and streaming at the same time.

Unlike traditional cable, DSL, or variable 5G wireless services that limit upload speeds, Empire Fiber Internet offers matching upload and download speeds to eliminate upload lag. This symmetrical capability ensures clearer video conferencing, faster file transfers, seamless photo and document backups, and reliable multi-device support for modern connected homes. Households can check service availability at their address to find an Empire Fiber Internet plan that matches their connectivity needs.

To read the original article, please click here.

The post Why Symmetrical Internet Speeds Matter for Your Home appeared first on Data Center POST.

TL;DR

• Expanding a Strategic Investment: Ares Secondaries funds have increased their investment in Sabey Data Center Properties (SDCP), bringing their total commitment to the platform to more than $500 million.
• Supporting Long-Term Growth: The additional investment builds on Ares’ initial minority equity investment announced in July 2026 and supports SDCP’s continued growth as demand for digital infrastructure increases.
• Scaling for Future Demand: With 275 MW across its existing campuses and an expansion pipeline expected to reach approximately 737 MW by 2033, SDCP is positioned to continue serving growing cloud and AI infrastructure needs.

# # #

As demand for digital infrastructure continues to grow, Sabey Data Center Properties (SDCP) is building on its long-term strategy with an expanded investment from Ares Secondaries funds.

The additional commitment follows Ares’ minority equity investment in SDCP announced in July 2026 and brings its total investment in the platform to more than $500 million. The follow-on investment further strengthens the relationship among Sabey Corporation, National Real Estate Advisors and Ares, bringing together three long-term partners focused on supporting SDCP’s continued growth.

SDCP currently owns and operates hyperscale and enterprise data center campuses in top-tier U.S. data center markets totaling 275 megawatts (MW). Looking ahead, the company’s expansion pipeline is expected to reach approximately 737 MW by 2033, positioning the platform to support the next generation of cloud and artificial intelligence demand.

“This follow-on investment reflects continued conviction in SDCP’s long-term value proposition and growth trajectory,” said Kevin Verdi, Executive Vice President & Chief Investment Officer at National Real Estate Advisors. “We remain focused on disciplined growth and delivering the digital infrastructure needed to support the next generation of cloud and AI demand.”

For SDCP, the expanded investment reinforces confidence in a platform built around developing and operating mission-critical infrastructure for customers across the United States.

“We are pleased to deepen our relationship with Ares as we continue executing on our long-term growth strategy,” said Tim Mirick, President of Sabey Data Center Properties. “This additional commitment reflects confidence in the strength of our platform and our ability to develop and operate the mission-critical infrastructure our customers rely on.”

The investment also reflects Ares’ continued confidence in SDCP’s growth potential.

“We are pleased to expand our investment in SDCP and build on our relationship with the Sabey and National teams as we continue to see significant runway for growth,” said Jamie Sunday, Co-Head of Real Estate Secondaries at Ares.

With a portfolio of more than four million square feet of mission-critical space, SDCP is one of the largest privately owned multi-tenant data center owners, developers and operators in the United States. As the digital economy continues to evolve, the expanded Ares investment provides additional support for SDCP’s long-term growth and its ability to serve customers with scalable, mission-critical data center infrastructure.

The post Sabey Data Center Properties Builds on Growth Strategy with Expanded Ares Investment appeared first on Data Center POST.

TL:DR

• Growing Grid Constraints: Expanding AI and digital workloads have made reliable electric grid access a major bottleneck for new data center capacity
• Repurposing Stranded Infrastructure: Utilizing orphaned and end-of-life oil and gas wells via in-situ hydrogen generation enables clean, low-cost fuel production near major power loads
• Zero-Loss Liquid Storage: Integrating onsite production with liquefaction and active zero-loss storage eliminates fuel boil-off and preserves energy reserves until needed
• Independent & Resilient Dispatch: Converting stored hydrogen through fuel cells provides clean, dispatchable power for backup reserves, peak demand management, or primary generation independent of electric grids

# # #

Data centers face a fundamental power challenge. As AI, cloud computing, and digital services drive electricity demand, access to reliable power is becoming one of the biggest constraints on new capacity. At the same time, operators are under growing pressure to reduce emissions, improve resilience, and limit dependence on increasingly constrained electric grids. The answer may be closer than many developers realize: thousands of existing oil and gas wells that are no longer economically viable.

Across the United States, documented orphaned oil and gas wells represent a significant underutilized infrastructure resource. Rather than viewing these wells solely as environmental liabilities, they could serve as platforms for producing hydrogen near where energy is needed. A new approach to in-situ hydrogen generation provides a path to do exactly that.

Turning stranded wells into hydrogen resources

Traditional hydrogen production requires bringing feedstocks and energy to a centralized facility, then transporting hydrogen to the end user. In-situ hydrogen generation turns that model on its head. The process uses existing end-of-life hydrocarbon reservoirs and well infrastructure. Oxygen is injected through existing or newly developed wells, triggering controlled downhole reactions that oxidize residual hydrocarbons and produce a hydrogen-rich syngas. At the surface, hydrogen can be separated and directed to the application that requires it.

This approach has the potential to produce hydrogen at very low cost while making productive use of existing subsurface resources and infrastructure. The process can also provide a pathway to lower-carbon hydrogen, with carbon management strategies determining the ultimate carbon intensity.

For data center developers, the significance is not simply how hydrogen is produced. It is where it can be produced. Locating hydrogen production near major loads can reduce reliance on long-distance fuel transport and provide a dedicated energy resource for power generation. In regions where suitable wells are near growing electricity demand, this could create an entirely different model for developing data center power infrastructure.

From hydrogen production to reliable power

Producing hydrogen onsite is only part of the opportunity. The hydrogen must also be available when the data center needs it. That makes storage critical.

Liquid hydrogen offers a highly energy-dense way to store substantial quantities of hydrogen onsite. Advanced zero-loss storage systems can actively remove heat from the tank, preventing hydrogen from warming and converting to gas. Instead of allowing boil-off to become a fuel loss, the stored hydrogen remains available for future use. That changes the economics and operational value of hydrogen storage.

For a data center, stored hydrogen can serve as a strategic energy reserve to meet extended power needs when grid power is unavailable or constrained. Hydrogen can be converted back to electricity via fuel cells, providing a clean, dispatchable power source without the combustion emissions associated with conventional diesel generation. The result is a potential closed-loop energy architecture: produce hydrogen at or near the site, liquefy it, store it without losses, and convert it to electricity when the data center requires additional or backup power.

A different approach to data center power

The conventional model assumes the grid will supply most electricity and that backup generators will protect against outages. That model becomes more difficult as data center loads grow and grid interconnection timelines lengthen.

Hydrogen offers another option: build power resilience around a fuel that can be produced, stored, and dispatched independently of the grid. The opportunity is particularly compelling when three technologies are considered together. In-situ hydrogen generation can unlock stranded subsurface resources. Liquefaction can make hydrogen practical for high-energy-density storage. Zero-loss liquid hydrogen storage can preserve that fuel until it is needed.

Together, these technologies could transform otherwise unproductive wells into assets supporting critical digital infrastructure. The broader opportunity extends beyond backup power. As hydrogen production and storage systems scale, the same infrastructure could support primary or supplemental power generation, helping data centers manage peak demand, improve energy resilience, and reduce reliance on constrained grid capacity.

Data centers need more than electricity. They need reliable power, fuel security, scalability, and predictable operating costs. The next generation of data center development may therefore require a different question. Rather than asking only where the grid can deliver more power, developers should also ask where existing energy resources can be converted into new, resilient power sources.

Orphaned wells may be part of the solution. By combining in-situ hydrogen generation with liquefaction and zero-loss liquid hydrogen storage, an overlooked energy resource could become a foundation for powering the digital infrastructure of the future.

# # #

About the Author

Cullen Hall is Vice President of Product Development at GenH2, a subsidiary of Path2 Hydrogen AG and technology leader in liquid hydrogen infrastructure systems, including Zero-Loss Controlled Storage and advanced hydrogen liquefaction. He has more than 20 years of experience in engineering, cryogenics, and hydrogen technologies, specializing in liquid hydrogen infrastructure. His career includes leadership roles with Linde’s Praxair, Plug Power, and several hydrogen companies throughout Asia, with expertise in cryopumps, compressors, dispensers, and hydrogen refueling systems. Learn more at genh2.com.

The post Powering Data Centers with Hydrogen from an Overlooked Energy Resource appeared first on Data Center POST.

TL;DR

• Accountability Beyond Consumption: Data centers must be judged by the physical systems they replace as well as what they consume, requiring a testable impact ledger tied to a defined, specific service
• Full-Lifecycle Burden Accounting: The ledger must record direct impacts, including embodied materials, electricity, cooling, water, backup power, and local grid constraints, allocated across the entire delivery network
• Verified Downstream Outcomes & Rebound: Claimed environmental benefits must be proven through measured downstream physical changes rather than predictive models, crediting participants conservatively while factoring in rebound usage

# # #

My previous article argued that a data centre should be judged by what it replaces as well as what it consumes. That principle is useful only if it changes decisions. An operator seeking planning consent, a customer buying capacity and an investor assessing a campus need a ledger that can be tested after the promise is made.

Start with a defined service, not a general claim that digital infrastructure makes society more efficient. A building optimization workload, for example, can be assessed against a specified portfolio, period, service level and physical outcome. The comparison must be the same buildings delivering a comparable service without that intervention. An empty field is not a credible baseline for an operating digital service; neither is an implausibly wasteful building.

Put the burdens on the record

The direct account should cover the site’s construction, embodied materials, electricity, cooling, water, backup generation, hardware, networks, replacement and end of life. It should also identify local constraints. A contracted renewable supply does not by itself establish that the demand causes no additional burden on a constrained grid. Water consumption needs its catchment and seasonal context. A single power efficiency ratio cannot answer either question.

The record should be granular enough to allocate a reasonable share to the workload being assessed. That allocation will be uncertain in a shared facility, so publish the method and a range. The account cannot quietly exclude the sensors, access networks, edge equipment or validation systems needed to deliver the claimed service.

Measure the physical change

The benefit account begins downstream. Did the controls reduce measured building energy use after adjusting for weather, occupancy and operating hours? Did leakage detection result in repaired pipes and less water lost? Did predictive maintenance extend useful equipment life, or merely produce alerts? Separate an action taken from a model’s forecast of what might have happened.

Credit must be shared conservatively. A saving in a building depends on the owner’s investment, commissioning, facilities team, sensors, software, connectivity and computing. It cannot be claimed in full by every participant. The customer may report a reduction in its own inventory, while the service provider separately describes its contribution to an avoided-impact assessment. These are different accounts and should remain labelled as such.

Rebound belongs on the same page. Lower operating cost can increase use. New computing capacity can create demand that was absent from the baseline. If plausible additional demand reverses the result, the claim has not passed the decision test.

Give evidence an expiry date

A sound submission records the functional unit, baseline, measurement period, source data, allocation method, uncertainty and independent assurance status. Physical sensors require identity, calibration and maintenance records. An immutable digital record may protect provenance after data entry; it cannot make a faulty reading true.

The baseline will also change. Buildings are retrofitted, grids decarbonise, regulations tighten and alternative services improve. A claimed benefit should be reviewed against current conditions rather than carried indefinitely from an early pilot. The TUTUM framework proposes a review at least every three years and expiry after five unless the baseline is independently re-established. Those periods are a proposed governance rule, not an external reporting standard.

Use the ledger as a gate

For each impact category, show the direct burden, attributable avoided impact and rebound separately. Carbon, water, materials and local infrastructure effects require their own units and their own decision. A favourable carbon result cannot compensate for an unacceptable water or community burden by arithmetic.

Before approval, the parties should agree who owns each measurement, who can challenge the baseline, which claims may be used publicly, and what happens if measured performance falls short. Procurement can require evidence access and periodic recalculation. Planning and investment committees can distinguish demonstrated benefits from scenarios and place conditions on the latter.

The conclusion may be positive, negative or unresolved. That is the point of an honest ledger. It gives the industry a route to show where computing delivers a verified physical improvement, while making it harder to use a distant possible saving to excuse a present local impact.

# # #

About the Author

Mark Burns is Executive Director of TUTUM Group and a chartered development and infrastructure leader. His work focuses on asset governance, programme assurance, sustainability and the relationship between physical infrastructure and digital capital.

The post Data Centres Need a Net System Impact Ledger appeared first on Data Center POST.

TL;DR

• AI is reshaping data center operations, driving higher rack densities, increased power requirements, and a greater demand for liquid-to-chip cooling expertise.
• Resilience depends on more than  redundant infrastructure, requiring proven processes, real-time monitoring, strong documentation, reliable partners, and well-trained operational teams.
• Datalec Precision Installations (DPI) is helping customers navigate  these challenges across the full data center lifecycle, from deployment and fit-outs to maintenance and specialized cooling support

# # #

In episode 70 of NEDAS Live!, host Ilissa Miller sat down with Scott Longhurst, Data Center Services Director at Datalec Precision Installations (DPI), to discuss the evolving operational demands of AI-ready data centers. Drawing on more than 35 years in mission-critical environments, Longhurst explores what resilience truly means as facilities become more dense, power-intensive, and complex.

For Longhurst, resilience extends far beyond redundancy. It requires proven processes, strong documentation, effective monitoring, dependable supply chains, and well-trained teams that can respond intelligently as infrastructure and customer requirements change. However, the industry faces a growing workforce challenge as data centers are being built faster than skilled operations professionals can be trained.

Artificial intelligence is accelerating the shift toward higher-density deployments and liquid-to-chip cooling. Longhurst notes that rack requirements are rising quickly, putting new pressure on facility power, cooling, structural capacity, maintenance practices, and deployment timelines. Liquid cooling also introduces a new operational boundary between landlord-managed and customer-managed systems, making coordinated monitoring and clear accountability essential.

Proactive maintenance was another key focus of the conversation. Longhurst explains how a maintenance program should not simply follow a calendar. Operators need to use predictive monitoring, equipment trend data, lifecycle planning, and root-cause analysis to identify and address issues before they disrupt service. The approach he outlines can reduce risk, support customer commitments, and guide smarter investment in aging infrastructure.

The two also discussed the industry’s continued focus on efficiency. Higher allowable operating temperatures, more efficient UPS systems, variable-speed equipment, and updated cooling strategies are helping operators lower energy use while maintaining reliable performance. For DPI, the opportunity lies in supporting customers across the data center lifecycle, from fit-outs and modular deployments to commissioning, maintenance, and specialized AI cooling services.

Tune in to episode 71 of NEDAS Live! for the full discussion on how resilient operations, skilled teams, and proactive maintenance strategies will shape the data centers powering the AI era.

The post Building Resilient Data Centers for the AI Era with Scott Longhurst appeared first on Data Center POST.

TL;DR

• Shift to AI-Ready Infrastructure: Rising AI workloads are driving rack densities up to 227kW, requiring early alignment of power distribution, cooling, and connectivity.
• Execution and Repeatability: Navigating diverse APAC markets demands standardized deployment strategies paired with local regulatory and operational expertise.
• Quality Safeguards Speed: Rigorous installation, structured cabling, and comprehensive system testing prevent post-launch operational issues as facilities pre-lease rapidly.

 # # #

Written by iMiller Public Relations on behalf of Datalec Precision Installations (DPI).

Asia-Pacific’s data centre market is experiencing continued expansion, with its development pipeline reaching a record 26.5 GW in the first half of 2026 after growing by 7.1 GW in six months. However, roughly 82 percent of this total remains in the planning phase, compared to 4.8 GW actively under construction. As colocation vacancy rates fall and pre-leasing accelerates across the region, the industry’s primary bottleneck has shifted from securing demand to executing construction and delivery effectively.

Artificial intelligence is reshaping infrastructure requirements, placing greater power and thermal demands on facility design. High-density hyperscale deployments now reach rack densities as high as 227kW, requiring early coordination across power distribution, cooling, connectivity, and modular infrastructure. To navigate complex regional dynamics across different APAC markets, operators must combine global delivery standards with local market intelligence, utilizing standardized deployment procedures to ensure consistent quality and scalability.

Speed to market cannot compromise operational reliability. Maintaining strict quality controls during equipment installation, cable management, labeling, and integrated system testing protects long-term facility performance. Ultimately, the region’s digital ecosystem will be defined not by the sheer size of its planned megawatts, but by its capability to convert complex development pipelines into resilient, AI-ready operational capacity.

To read the original article, please click here.

The post From Pipeline to Performance: How APAC Data Centres Can Deliver AI-Ready Capacity at Scale appeared first on Data Center POST.

TL;DR

• Power as the Primary Bottleneck: Driven by AI and cloud expansion, digital infrastructure demand is at an all-time high, but grid capacity, interconnection timelines, and power availability now determine project timing and success.
• Practitioner-Led Advisory & Truth-Telling: Founded by former operators from Digital Realty, CoreSite, and Equinix, ASG pairs hands-on development experience with deep grid and regulatory expertise to deliver honest early feasibility assessments.
• Convergence of Power & Digital Infrastructure: Over the next decade, ASG foresees power and digital infrastructure merging into a single integrated asset class, where data centers are planned directly alongside energy generation and transmission assets

# # #

Data Center POST had the opportunity to connect with Jarrett Appleby, CEO & Managing Partner at ASG. Jarrett has spent his career on every side of the table: from running operations at Digital Realty and CoreSite, building go-to-market at Equinix, and leading strategy at global network operators like FLAG Telecom and Reliance Globalcom. When he founded Appleby Strategy Group, Jarrett wanted to give developers and investors the practitioner’s view that he would have wanted as an operator. That became ASG, a team of senior executives who have been in the trenches of data center, power and fiber development.

What is ASG’s core mission?

ASG helps clients turn digital infrastructure projects from concept into assets that are financeable and ready to transact. Power, land, permitting and capital now decide which projects succeed, and many projects stall between strategy and delivery. We bring strategy and execution under one roof. We also tell clients the truth about which projects are feasible and worth doing.

How would you describe the current state of the global digital infrastructure market in terms of challenges and opportunities for your business?

Demand has never been stronger, driven by AI and cloud. The main limit is now power, not demand. Interconnection timelines, grid capacity, permitting and supply chain decide who can deliver and when. That’s a challenge for the industry. It’s also our biggest opportunity: clients who secure power and site readiness early set themselves apart.

Who do you see as your primary customer base today, and how has that evolved over the past few years?

We work with data centers, power companies, developers, neoclouds, operators, PE/VC investors and hyperscalers. A few years ago, most engagements centered on traditional colocation and enterprise. Today, more of our work involves large AI campuses, new neocloud platforms, and investors entering digital infrastructure who need operating expertise.

What are the top priorities or concerns you hear from your customers, and how is your company addressing them?

• Speed to power: We provide power and interconnection advisory, including grid planning, queue strategy and cost allocation.
• Site certainty: We support site selection, land control, permitting and transaction readiness.
• Finding the right buyer or partner: We help identify buyers, set valuations and close deals.

Are there emerging customer segments or industries that you see as growth areas for your business?

AI and neocloud operators, energy and utility companies monetizing land and generation, and institutional capital entering the sector.

How are you aligning your company’s strategy to address sustainability and regulatory requirements in the coming years?

We build grid, energy-market and regulatory strategy into site development from day one. That includes interconnection cost allocation, on-site and hybrid generation, and working with utilities and regulators. Projects that are good grid citizens get approved and financed faster.

How is ASG fostering partnerships to create value for customers and enhance your competitive position?

We connect utilities, landowners, hyperscalers, operators and capital providers. Our team’s relationships across the ecosystem let us put the right counterparties together early, which lowers risk and shortens time to close.

What are ASG’s core products or services?

Strategic development and execution; site development strategy and leadership; go-to-market and partner facilitation; power and interconnection advisory.

What markets do you offer ASG’s solutions to?

We offer solutions to the global market. ASG is headquartered in Northern Virginia and advises on data center, power and fiber projects across North America and international markets.

What long-term goals or initiatives are you most excited about, and how do you envision the industry evolving over the next decade?

Over the next ten years, power and digital infrastructure will merge into a single asset class. Data centers will be planned alongside generation and transmission. We want ASG to be the go-to partner for that integrated development.

What upcoming industry events will you be attending?  

Over the next six months, ASG will be attending DCD Connect VA, Neocloud Summit (an invite- only ASG Event), PTC Hawaii, and CERAWeek by S&P Global.

Is there anything else you would like our readers to know about your company and capabilities? 

ASG was built by practitioners, for practitioners. Our leaders have run operations, raised capital, developed sites and negotiated with utilities, so we know what it takes to get projects built. That depth runs across disciplines: our team pairs data center operating experience with grid and regulatory expertise, including Allison Clements, a former FERC Commissioner. That combination matters more than ever now that power, not demand, sets the pace of development.

We also hold ourselves to a simple standard: we only take on work we believe is feasible and worth doing, and we tell clients the truth about it early, before time and capital are spent. Whether a client is a developer securing power and land, an investor underwriting a platform, or a hyperscaler or neocloud looking for capacity, we bring strategy and execution together to get projects to a financeable, transaction-ready outcome.

Strategic Partners. Industry Builders. Truth Tellers.

# # #

About ASG

ASG is the outcome-driven advisory partner behind leading data center, power and fiber projects. Founded by practitioners who have been developers, investors, operators and advisors, ASG provides strategic development, site development, go-to-market and power and interconnection advisory to developers, neoclouds, operators, investors and hyperscalers. ASG is headquartered in Northern Virginia. Learn more at asg-team.com or reach out to asg@asg-team.com.

The post Executive Profile: Jarrett Appleby on Power, AI, and Digital Infrastructure appeared first on Data Center POST.

TL;DR

• Rightfiber has completed its acquisition of Fastwyre Broadband’s Nebraska business, adding 26 communities to its footprint.
• Former Fastwyre customers and employees now join a provider serving 20 states, with more than 28,000 route miles of fiber and over 300,000 FTTH passings.
• Integration has begun, and customers will receive ongoing updates on products, services, billing, and future enhancements.

# # #

For more than a century, Nebraska has been an important part of Rightfiber’s story. Now, the fiber provider is expanding that presence as it enters its next chapter of growth.

Rightfiber, formerly Great Plains Communications and a portfolio company of Grain Management, has completed its acquisition of Fastwyre Broadband’s Nebraska business, bringing 26 additional communities across the state into its growing network.

With the deal done, former Fastwyre customers in Nebraska are now part of one of the nation’s largest independent fiber broadband providers. Rightfiber’s network spans more than 28,000 route miles of fiber across 20 states and supports over 300,000 Fiber-to-the-Home (FTTH) passings.

For customers, the focus now shifts to the transition. Rightfiber has begun integrating Fastwyre’s Nebraska operations and will continue communicating with customers about products, services, billing and future enhancements.

“We’re excited to welcome Fastwyre’s Nebraska customers and employees to Rightfiber,” said Heath Simpson, Chief Executive Officer of Rightfiber. “We look forward to serving these communities with reliable connectivity, responsive local support, and the customer-first experience that define Rightfiber. Our focus is on making this transition as seamless as possible.”

The acquisition also builds on a history in Nebraska that stretches back more than 115 years. While the Rightfiber name is new, the company’s connection to the state is not.

“Nebraska has been an important part of our story for more than 115 years, and completing this acquisition reinforces our long-term commitment to the communities we serve across the state,” said Todd Foje, Executive Chairman of Rightfiber. “We’re excited to build on that legacy by strengthening our relationships across Nebraska and continuing to support the people, businesses, and communities that have been the foundation of our success for generations.”

That combination of history and growth helps put the acquisition into perspective. Rightfiber is expanding the scale of its network while building on an established presence in Nebraska. As Rightfiber continues its next chapter, the Fastwyre acquisition represents another step in that evolution: expanding the network, welcoming new communities. and carrying its Nebraska legacy forward under the Rightfiber name.

Read the full release here: www.imillerpr.com/news/rightfiber-completes-acquisition-of-fastwyre-broadbands-nebraska-business.

The post Rightfiber Builds on Its Nebraska Roots With Continued Fiber Expansion appeared first on Data Center POST.

TL;DR

• Fuel readiness depends on more than having enough fuel on-site. Operators need visibility into how fuel is performing and how quickly it is being consumed.
• Real-time fuel data can help teams compare expected runtime with actual conditions and make more informed replenishment and emergency planning decisions.
• Operational intelligence can help identify developing fuel-quality or system-performance issues before they affect generator readiness.
• By connecting fuel condition, consumption, inventory, and system performance across the lifecycle, organizations can reduce uncertainty and make more confident operational decisions.

# # #

For decades, fuel has often been treated as a relatively straightforward part of backup power planning. As long as there is enough fuel on-site and the system is maintained, backup power is expected to be ready when needed.

For mission-critical operations, however, fuel readiness requires a clearer understanding of what is actually happening across the system. Operators need visibility into fuel performance and consumption so they can better understand how long available supply will support operations and whether changing conditions could affect readiness.

This is where operational intelligence becomes valuable. Instead of viewing fuel as a static asset sitting in a tank, organizations can use data from across the fuel lifecycle to better understand system performance and make more informed decisions before an outage puts those assumptions to the test. ESI Total Fuel Management’s approach reflects this shift by treating fuel as an actively managed operational variable.

From Fuel Data to Operational Intelligence

Collecting data is only useful when that information can help guide decisions. A fuel system may provide visibility into inventory, consumption, and equipment performance, but those data points become more valuable when they are viewed together and used to understand overall readiness.

Knowing how much fuel is stored on-site, for example, does not necessarily tell operators how long that supply will last. Actual consumption can change under load, and runtime assumptions may not always reflect real operating conditions. ESI’s fuel consumption monitoring capabilities help provide that added visibility by tracking engine supply and return flow rates in real time.

Moving Beyond Runtime Assumptions

Backup power planning often starts with expected runtime, typically based on tank capacity and estimated generator consumption. While those calculations are useful, actual operating conditions can change how long available fuel will support the facility.

Consumption may shift as generator load changes, and fuel may be distributed across different parts of the system. During an extended outage, those variables can make planned runtime different from actual runtime.

Operational intelligence helps give operators a clearer picture of what is happening in real time. By comparing current fuel consumption with available inventory, teams can make more informed decisions about replenishment and emergency fuel needs before supply becomes a concern.

That visibility is especially important during prolonged outages or other high-stress events, when understanding actual runtime can directly support continuity planning. ESI’s approach is designed to help mission-critical operators move from estimated runtime to a more informed view of how long their systems can actually perform.

Seeing Risk Before It Becomes a Failure

Fuel intelligence goes beyond knowing how much fuel is available. Fuel condition and overall system performance also affect whether backup power will perform as expected when it is needed.

In standby applications, fuel may remain stored for long periods, creating opportunities for contamination or other issues to develop over time. Scheduled inspections and testing can help identify problems, but an intelligence-driven approach provides a more continuous view of system conditions so emerging risks can be addressed earlier.

ESI’s model is built around that broader view of readiness. Instead of evaluating fuel quality or equipment performance in isolation, ESI treats the fuel system as an interconnected whole and emphasizes continuous verification. This gives operators greater visibility into developing concerns before they have the chance to affect generator performance.

Turning Visibility Into Better Decisions

Operational intelligence is most valuable when it helps teams take action with greater confidence. Visibility into fuel condition, consumption, inventory, and system performance gives operators a stronger foundation for decisions across the fuel lifecycle.

Operational intelligence can also improve lifecycle efficiency by giving teams better information about how fuel is being used and where attention may be needed. This can support more informed maintenance decisions, help extend fuel life, and reduce unnecessary waste without compromising readiness.

During an extended outage, current operating information becomes even more important. Facility teams and fuel partners can use that information to coordinate replenishment and response based on actual conditions rather than relying only on predetermined plans.

Across each of these areas, the goal is the same: Reduce uncertainty and give operators clearer, more timely information to support better decisions.

Connecting Intelligence Across the Fuel Lifecycle

A single data point cannot provide a complete picture of fuel readiness because each part of the fuel lifecycle is connected. From sourcing and delivery through storage, conditioning, transfer, consumption, and replenishment, conditions in one area can influence performance in another.

That is why operational intelligence becomes more valuable when it is part of a broader fuel management strategy rather than treated as a standalone monitoring function. ESI brings together engineered systems, lifecycle services, operational intelligence, fuel access, and emergency readiness to manage fuel as one integrated system.

This creates an important distinction between simply seeing what is happening and understanding what to do next. Monitoring can provide visibility into current conditions, while operational intelligence helps operators interpret that information in context and use it to support decisions across the fuel lifecycle.

From Fuel Visibility to Operational Confidence

Mission-critical operators do not need more data simply for the sake of having it. They need information that gives them a clearer understanding of current conditions, helps identify what may be changing, and supports timely action when it is needed.

That is the value of actionable fuel intelligence. When fuel condition, consumption, inventory, and system performance are viewed together, teams can make decisions based on a more complete picture of readiness rather than relying on assumptions.

For ESI, this means managing fuel as more than something stored on-site for an emergency. Fuel is treated as a performance-critical asset that can be monitored and actively managed throughout its lifecycle, giving operators greater visibility and confidence in how the system is performing.

When backup power is essential to uptime, better information can support better preparation and more informed decisions before an issue has the chance to affect operations.

To learn how ESI Total Fuel Management can help turn fuel-system data into actionable operational intelligence, contact ESI Total Fuel Management today.

The post Transforming Fuel into Actionable Intelligence for Smarter Decisions appeared first on Data Center POST.

TL;DR

• Space Is Not the Constraint, Usable Space Is: A 5 MW requirement for rack-scale GPU systems went to more than ten operators. Nearly all had space; almost none could support the density on the timeline
• Liquid-Cooling Ready Is Not Liquid-Cooled: The date that matters is when facility cooling reaches the white space, and it often sits well behind the building
• Credit Can Matter as Much as Capacity: Two operators raised, unprompted, whether the tenant was investment grade
• Buyers Have More Flexibility Than They Use: Phasing, multiple sites, air-cooled platforms and an earlier start all widen the field

# # #

By James Mercer, Principal, Metro Colo Advisory

North American data center vacancy has never been lower. CBRE put it at a record 1.4 percent across primary markets in the first half of 2026. It is easy to read that as “there is no space.” For AI buyers, the more accurate reading is harder: there is space, but very little of it can run what they are trying to deploy.

This year we took a requirement to market that tested that distinction directly: 5 MW of capacity for rack-scale GPU systems, liquid-cooled, commissioning in early 2027, with a path to roughly twice that within a year. We took it to more than ten operators across North America. Nearly every operator had space. Almost none could support the requirement. Here is why, and what buyers can do about it.

The Gap Between Space and Usable Space

Current rack-scale systems such as NVIDIA’s GB300 NVL72 draw 140 to 160 kW per rack. Most of the market was not built for that. According to the Uptime Institute, rack density growth is concentrated between 10 and 30 kW, about one facility in eight reports any racks at 30 to 59 kW, and racks above 100 kW remain rare.

Three things decide whether a hall can take this hardware, and none of them can be changed quickly:

Power density per rack. A hall designed for 15 kW racks cannot be averaged up to 150. The electrical distribution to each position has to exist.

Where the cooling loop ends. Direct liquid cooling needs facility water delivered to the rack row. A building with chilled water at the plant room but not in the white space is not a liquid-cooled building.

Floor loading. A fully populated rack-scale system weighs about 3,000 pounds. Many raised floors were never rated for it.

An operator can have megawatts of vacant, powered space and still have nothing that clears all three.

“Liquid-Cooling Ready” Is Not Liquid-Cooled

The most common answer we heard was a variation of “our facilities are liquid-cooling ready.” In practice that meant the building could support liquid cooling once equipment was procured and installed, and one operator put the lead time for facility-side cooling equipment at nine to twelve months. In one case, a building was due in early 2027 while its liquid-cooled space was not due until months later.

It also helps to separate two kinds of equipment that often share a name. The coolant distribution units inside the racks usually ship with rack-scale systems. The facility-side units that connect the building’s water to the rows belong to the operator. When an operator cites a long lead time, it is worth asking which of the two they mean.

The Gate Nobody Mentions: Credit

Two of the operators that engaged most seriously raised the same question, unprompted: was the tenant investment grade? One made a meeting about its future capacity conditional on the answer.

That makes sense from the operator’s side. A multi-year, multi-megawatt commitment is a large receivable, and in a market this tight an operator can choose its tenants. But it means some declines framed as “no capacity” may be credit decisions. AI buyers without investment-grade credit should expect the question, and arrive with an answer: a parent guarantee, a deposit, or a stronger entity on the lease.

What Buyers Should Ask Before Shortlisting a Facility

1. What density can you support per rack, sustained, in the hall you are offering? Not the campus maximum.
2. On what date will facility cooling be live at the rack row? Not the building handover date.
3. What is the floor rating in that hall?
4. What is the expansion path, and is it approved or on a roadmap? Many operators can start a deployment. Far fewer can commit to doubling it.
5. What will you need to see on credit? Asking early saves weeks.

Where the Flexibility Is

Buyers have more room than the headline requirement suggests. Starting smaller with a committed expansion path opens more facilities than insisting on the full load on day one. Splitting a deployment across two sites works when the workloads are independent. And where the workload allows, air-cooled GPU servers at 30 to 50 kW per rack fit a far larger share of the market than rack-scale systems do.

The biggest lever is time. Capacity for 2027 is being committed now: CBRE found 80.4 percent of capacity under construction already preleased. A buyer who starts twelve to eighteen months ahead, with credit support lined up and flexibility on phasing, is choosing from a market. A buyer who starts six months ahead is choosing from whatever is left.

# # #

About the Author

James Mercer is Principal of Metro Colo Advisory, an independent data center advisory practice based in New York. He places colocation, wholesale and AI infrastructure requirements with operators across North America, and advises developers and investors on how facilities will compete for tenants.

The post Vacant Is Not Usable: What a 5 MW AI Capacity Search Found appeared first on Data Center POST.

TL;DR

• Following its 2026 relaunch, IGF-USA is exploring how it can evolve beyond an annual conference into a sustainable, year-round organization.
• Building institutional memory and separating administrative operations from substantive programming could help IGF-USA maintain continuity while bringing new participants into the organization.
• A decentralized national model could expand Internet governance discussions beyond Washington and involve universities, state and local governments, Tribal nations, businesses and communities.
• Youth and academic engagement could bring the next generation directly into Internet governance through student programming, curriculum integration and active participation.
• Digital infrastructure issues including land use, energy, community engagement and infrastructure development are increasingly part of Internet governance and demonstrate the need for broader participation.

# # #

By Ilissa Miller, CEO of iMiller Public Relations

One of the most interesting parts of IGF-USA 2026 happened after our data center discussion. The final session turned the governance conversation inward.

The question became, after relaunching IGF-USA following a multi-year hiatus, what should the organization itself become?

It’s a timely question. The global Internet Governance Forum now has a permanent mandate from the United Nations. Meanwhile, AI, cybersecurity, digital infrastructure, privacy and other technology issues are evolving rapidly, and increasingly playing out far beyond Washington.

Learning From the Past

Christian Dawson brought important institutional history to the conversation. Nearly a decade ago, he co-chaired an IGF-USA Sustainability Working Group focused on creating processes and governance structures that could make the organization sustainable.

One of the lessons from that experience was the importance of institutional memory.

Organizations built heavily around volunteers can become dependent upon a relatively small number of people who know how everything works. When those people move on, knowledge can disappear with them.

Christian captured the challenge beautifully: IGF-USA needs “continuity without dependency and renewal without exhaustion.”

That’s not only a useful principle for IGF-USA. It’s a pretty good governance principle for almost any multistakeholder initiative.

Separate Operations From Ideas

Ron da Silva, Chair of the IGF-USA 2026 Steering Committee, discussed another practical idea: separating the administrative responsibilities of running IGF-USA from the substantive programming.

A steering committee can handle the mechanics necessary to keep the organization functioning while a broader programming community focuses on the issues, participants and conversations that matter.

Leadership can rotate. Institutional knowledge can be documented. New people can participate without requiring the organization to reinvent itself every year.

And, I think the bigger idea emerged as the discussion continued.

Don’t Just Move the Conference. Move the Conversation.

If IGF-USA is truly a national Internet Governance Forum, why should its center of gravity remain primarily in Washington? Internet governance is happening everywhere.

It’s happening in state legislatures. Universities. Tribal nations. Planning boards. Utilities. Technology companies. Economic development organizations. And, as our earlier panel demonstrated, communities deciding whether and how digital infrastructure should be developed.

The idea discussed was not simply rotating an annual conference around the country. It was creating a decentralized national network.

Changing geography should change who participates, which institutions become involved and, importantly, which questions get asked.

That could mean universities helping shape programming. Local businesses bringing regional technology issues forward. State and local officials participating directly. Civil society organizations introducing concerns that might never surface in Washington.

Bring the Next Generation With Us

Chelsea Horn of Lehigh University added another critical component: youth and academic engagement.

Her ideas included academic affiliates, student scholars, youth-led programming, curriculum integration and involving students directly as moderators, rapporteurs and participants.

If we’re discussing how the Internet and AI should evolve over the next several decades, the people who will inherit those systems shouldn’t simply be invited to watch the conversation.

The Data Center Panel Proved the Point

What I found particularly interesting was that the data center session I was part of earlier in the day became something of a proof point for this proposed model.

During the final discussion, organizers specifically noted that the panel brought people and perspectives into the IGF conversation that aren’t normally part of the traditional Washington Internet-policy world: Local planning; Tribal sovereignty; Community engagement; Land use. Energy; and  Infrastructure development.

Those aren’t separate from Internet governance anymore. They are Internet governance.

And that’s why I left Philadelphia convinced that IGF-USA shouldn’t simply be an annual event.

It has an opportunity to become a decentralized, year-round ecosystem where industry, government, academia, technical experts and communities continually work through the difficult questions accompanying our digital future.

The Internet itself is decentralized.

Perhaps the conversation about how we govern it should be too.

And for those of us working across digital infrastructure, there is every reason to participate. Because the next generation of Internet governance isn’t only about what happens online. Increasingly, it’s about what happens in our communities.

Curious about what you missed? Not to worry, you can still view the live stream online here: isoc.live/21744.

Stay tuned for IGF-USA upcoming events: igfusa.us.

The post Beyond the Annual Conference: What’s Next for IGF-USA? appeared first on Data Center POST.

TL;DR

• Data Fragmentation Challenge: The data center industry suffers from widespread data fragmentation, with critical information scattered across energy markets, land records, regulatory systems, supply chains, capital markets, and sustainability frameworks.
• Sovereign Intelligence Layer: Rather than merely storing isolated datasets, organizations need a sovereign intelligence layer to connect analytical silos into a governed, decision-ready environment.
• Critical Role of Data Provenance: As AI accelerates analytical speed, maintaining verifiable source evidence and traceable data provenance is essential for high-stakes, multimillion-dollar capital decisions.
• Data as Decision Infrastructure: Market advantage will belong to organizations that move beyond data accumulation to build connected, scalable decision infrastructure.

# # #

The strategic case for connecting fragmented infrastructure data into governed, decision-ready intelligence. 

The data center industry does not suffer from a shortage of data. It suffers from fragmentation. 

Operators, developers, investors, infrastructure providers, and technology companies increasingly make decisions using information scattered across energy markets, land records, regulatory systems, supply chains, capital markets, sustainability frameworks, and geospatial databases. 

Each dataset may be useful independently. The larger challenge is turning them into a coherent picture quickly enough to support consequential infrastructure decisions. 

As artificial intelligence accelerates demand for digital infrastructure, this distinction becomes increasingly important. The next generation of data infrastructure should not merely store information. It should help organizations understand the relationships between information, establish its provenance, and convert it into decision-ready intelligence. 

THE FRAGMENTATION PROBLEM 

Consider a hypothetical data center development. 

Before construction begins, decision-makers may need to understand land ownership, zoning, power availability, grid constraints, water resources, fiber connectivity, environmental requirements, financing conditions, construction costs, supply-chain exposure, and local regulations. 

Those questions frequently reside in different systems maintained by different institutions. 

Traditional analytics can answer individual questions. Decision infrastructure must address the relationships among them. 

A parcel of land may appear attractive until energy constraints are incorporated. An energy source may appear economical until regulatory requirements are considered. A development opportunity may appear financially sound until infrastructure costs or supply-chain dependencies are incorporated.

The value therefore lies not simply in possessing more information, but in connecting information that previously existed in separate analytical silos. 

AI IS ONLY PART OF THE ARCHITECTURE 

Artificial intelligence offers extraordinary capabilities for analyzing large amounts of information, but AI does not eliminate the need for reliable source data. 

An intelligence system is only as useful as its underlying evidence. 

Organizations deploying AI-assisted decision systems should therefore consider several questions: Where did the information originate? When was it collected? Which jurisdiction does it concern? Can the underlying source be verified? Has the information changed? What assumptions were introduced during analysis? 

These questions become particularly important when analytical outputs influence capital allocation, infrastructure development, or regulatory decisions. 

The objective should be an architecture in which AI accelerates analysis while preserving a clear path back to the underlying evidence. 

TOWARD SOVEREIGN DECISION INFRASTRUCTURE 

A useful concept emerging from this environment is the sovereign intelligence layer. 

Here, sovereignty does not necessarily mean isolation. It means maintaining control over how information is sourced, governed, interpreted, and used. 

For data-center organizations, such an intelligence layer could combine several domains: infrastructure and energy intelligence; property and geospatial information; regulatory and jurisdictional data; supply-chain information; corporate and counterparty intelligence; financial and capital-market information; and sustainability and environmental indicators. 

The resulting system becomes more than a dashboard. 

It becomes an analytical environment capable of showing how changes in one domain affect decisions elsewhere. 

PROVENANCE BECOMES INFRASTRUCTURE 

As AI-generated analysis becomes commonplace, provenance may become as important as processing power. 

Decision-makers need to distinguish among verified facts, estimates, assumptions, historical information, and machine-generated interpretations. 

That requires systems designed around evidence rather than merely outputs. 

Every significant analytical conclusion should ideally have a traceable path back to its underlying information.

This principle becomes especially important in infrastructure environments where a seemingly small error can influence multimillion-dollar capital decisions. 

WHAT COMES NEXT 

The rapid expansion of AI infrastructure will create enormous quantities of operational and market information. The organizations that gain the greatest advantage may not simply be those possessing the largest datasets. 

They may be the organizations capable of connecting disparate datasets while maintaining provenance, governance, and context. 

That requires moving beyond the traditional idea of data as something organizations merely collect. 

Data must increasingly become decision infrastructure. 

The future data center will therefore be shaped not only by processors, power systems, cooling technology, and connectivity. It will also depend upon the intelligence architecture surrounding those physical systems. 

As digital infrastructure becomes more interconnected with energy, capital, property, regulation, and public policy, the ability to see those relationships clearly could become one of the industry’s most valuable capabilities. 

# # #

About the Author

Safwan Bey is Chairman and CEO of Trinlah Sovereign Holdings Corporation (TSHC), where his work encompasses data intelligence, infrastructure, private capital, sustainability, and technology. He leads the development of the TSHC Sovereign Data Foundry, an initiative focused on transforming fragmented public, institutional, and commercial information into structured, decision-grade intelligence. 

The post From Data Overload To Decision Infrastructure: Why Data Centers Need A Sovereign Intelligence Layer appeared first on Data Center POST.

TL;DR

• Executive Appointment: Datalec Precision Installations (DPI) appointed Dominik Dziarczykowski as Business Development Manager to bolster its high-density and liquid cooling capabilities.
• Proven Track Record: Dziarczykowski brings more than 20 years of industry experience, including previous liquid cooling roles at CoolIT Systems and Vertiv. 
• AI & High-Density Focus: Joining Richard Dobbie’s Business Development team, he will focus on providing practical cooling solutions for AI-driven computing needs.

# # #

Written by iMiller Public Relations on Behalf of Datalec Precision Installations (DPI).

Datalec Precision Installations (DPI) has appointed Dominik Dziarczykowski as Business Development Manager to strengthen its capabilities in liquid cooling and high-density data center infrastructure. Joining the Business Development team under Sales and Marketing Director Richard Dobbie, Dziarczykowski will support customers and partners as DPI continues to expand its direct liquid cooling solutions and associated services. 

Dziarczykowski brings over 20 years of experience across the IT, consulting, and data center industries. His professional background includes roles at major global organizations such as IBM, Dell, Huawei, Deloitte, and Capgemini, along with direct liquid cooling leadership at CoolIT Systems and Vertiv. In his new position, he will assist clients with practical liquid cooling deployments, debuting as a company representative with a fireside chat at Data Centre World Asia 2026 in Singapore. 

Datalec’s leadership highlighted that his technical and market expertise aligns directly with the company’s mission to deliver end-to-end, future-ready infrastructure capable of supporting high-density workloads and the AI revolution. 

To read the full press release, please click here.

The post Datalec Strengthens its Liquid Cooling Capabilities with the Appointment of Dominik Dziarczykowski appeared first on Data Center POST.

TL;DR

• Shift Focus from Deployment to Long-Term Operations: While the urgency to bring AI capacity online drives rapid installation, long-term success depends on treating liquid cooling as an ongoing operational discipline rather than just a quick procurement project.
• Holistic Resilience & Blast Radius Control: Evaluating system resilience requires looking beyond redundant Cooling Distribution Units (CDUs) to minimize the overall failure “blast radius” across the entire fluid distribution and control network.
• Proactive Performance Management & Adaptability: Leveraging real-time operational data allows operators to catch performance drift before component failure occurs, ensuring cooling systems continuously adapt as AI hardware and dynamic workloads evolve.

# # #

by Kevin Roof, Director of Offer & Capture Management, LiquidStack

Much of today’s discussion around liquid cooling centers on deployment. Which CDU should I buy? Which cooling technology should I choose? What rack density can I support? These are important questions, but they represent only the beginning of the liquid cooling journey.

There is an understandable urgency to bring AI capacity online as quickly as possible. But in some cases, the pace of deployment is beginning to outstrip the operational maturity surrounding liquid cooling. Many organizations are implementing these systems for the first time, often without the decades of established practices that exist around more traditional data center infrastructure.

As AI infrastructure grows more critical the question therefore needs to shift from “How do we deploy liquid cooling?” to “How do we keep these systems operating efficiently for the next decade?”

Installation is measured in weeks. Operations are measured in years.

Liquid Cooling Operations Are an Engineering Discipline

Many operators still approach liquid cooling primarily as a procurement and deployment project: find the right CDU, integrate it into the environment, bring the GPU clusters online and move on to the next capacity challenge. But liquid cooling systems are infrastructure platforms requiring careful and continuous management.

Design establishes the conditions for long-term performance. Commissioning validates that those decisions work in practice. Operations determine whether that performance can be sustained as workloads, hardware and operating conditions change.

That last point is increasingly important. Traditional data center operations were largely built around maintaining equipment. With liquid cooling, the technology cooling system, facility infrastructure, controls and IT equipment operate as a highly interconnected system. Changes to rack configurations, accelerators, control strategies or even routine maintenance can affect behavior elsewhere in the cooling loop.

That is why liquid cooling operations increasingly need to be viewed as an engineering discipline rather than simply a maintenance function.

Fluid Management Means More Than Filling the Loop

One of the biggest misconceptions surrounding liquid cooling is that coolant is simply another utility, much like chilled water flowing through a facility.

In reality, fluid management is one of the most important determinants of long-term system reliability. The fluids circulating through a liquid cooling system must remain compatible with every material they encounter, from piping and seals to cold plates and heat exchangers.

Maintaining proper chemical balance helps minimize corrosion, prevent scaling, reduce biological growth and preserve consistent thermal performance. Small deviations may not create immediate problems, but over time they can gradually sap efficiency and shorten the lifespan of critical components. Internal channels on advanced coldplates can be as small as 100µm, which is about the diameter of a human hair. Even minor biological growth or debris from scaling can easily clog coldplate channels or the filters in place to protect them. 

The important mindset shift, then, is to stop treating fluid quality as something checked only during scheduled service. It is an ongoing operational condition that can affect virtually every component in the system.

Resilience Is About More Than Redundant CDUs

High availability has always been a cornerstone of data center design, but redundancy becomes even more important as AI workloads grow in value and computational intensity.

Most operators understand concepts such as N, N+1 or N+2 configurations. Similar principles apply to liquid cooling, but redundancy should extend well beyond the CDU itself.

One mistake is evaluating resilience too narrowly at the component level. A redundant CDU does not necessarily create a resilient cooling architecture if another shared element can interrupt cooling across an entire compute pod.

A better way to think about resilience is in terms of blast radius: if something does fail, how much of the environment does that failure affect?

A resilient system considers pumps, heat exchangers, sensors, valves, controls, the broader fluid distribution network, as well as the upstream power architecture and TCS loop layout. The goal is not simply to eliminate individual points of failure. It is to contain failures and ensure that maintenance or component replacement can occur without unnecessarily disrupting compute operations.

Maintenance Is Becoming Performance Management

Liquid cooling also creates an opportunity that operators should take advantage of: these systems generate a wealth of operational data.

Flow rates, pressure differentials, supply and return temperatures, pump performance, fluid quality and leak detection collectively provide a detailed view into overall system health. The question is whether organizations use that information simply to react to alarms or to understand how system performance is changing over time.

Preventive maintenance has traditionally centered on fixed intervals. Fluid is sampled, filters are inspected, pumps are verified, sensors are calibrated and valves are tested according to a schedule. Those activities remain essential, but they no longer tell the whole story.

In increasingly dynamic AI environments, understanding how the system behaves between maintenance intervals can be just as important.

Small deviations in pressure, pump current, temperature or flow stability may not indicate an immediate problem. But they can provide early evidence that conditions are drifting. By the time a component fails, the system may have been signaling deterioration for weeks.

That turns maintenance into performance management. Instead of waiting for failures, operators can identify trends, investigate abnormalities and schedule service before small changes become production issues.

There is an organizational dimension here as well. Liquid cooling increasingly blurs the boundaries between facilities, IT teams, tenants and equipment vendors. If different stakeholders can make changes that influence flow, pressure or thermal behavior, organizations need clear ownership around who monitors the system, who can change parameters and how visibility is shared.

Operational resilience increasingly depends on coordination as much as hardware.

The System You Install Is Not the System You Will Operate

Perhaps the biggest mistake organizations can make is designing and operating exclusively for today’s realities.

AI infrastructure is evolving at an extraordinary pace. Cooling systems deployed today may be expected to support several generations of accelerator technology, each bringing different thermal characteristics, flow requirements and operating behaviors. And it is not simply a question of supporting more heat. 

Training environments can introduce rapid swings in compute demand, requiring cooling systems to respond dynamically. Inference environments may create steadier, sustained loads. Hardware refreshes can change hydraulic requirements even when the underlying cooling infrastructure remains physically unchanged.

That means successful commissioning does not establish a permanent operating state.

When accelerators change, racks are reconfigured or significant maintenance alters the cooling loop, operators should reassess whether the assumptions under which the system was originally commissioned still apply. In some cases, changes may warrant new control strategies, different setpoints or even renewed system validation.

A well-operated liquid cooling system should not be defined by its ability to remain unchanged. It should be defined by its ability to adapt while continuing to deliver predictable thermal performance.

Installation Is Only the Beginning

Liquid cooling has become one of the foundational technologies enabling the AI era, but technology alone does not guarantee success.

The true measure of a deployment is not how quickly the equipment is installed or how well the system performs on commissioning day. It is whether that infrastructure continues to deliver reliable, efficient cooling years later as workloads expand, hardware changes and operating conditions evolve.

That requires a different mindset. Installation is not the finish line, and commissioning does not freeze a cooling system in time. Liquid cooling must be continuously observed, maintained, adjusted and, when conditions materially change, revalidated.

The industry has spent enormous energy learning how to deploy liquid cooling for AI. Now we need to become equally good at operating it.

# # #

About the Author

Kevin Roof is the Global Director of Offer & Capture Management for LiquidStack. A mechanical engineer and PMP with over a decade of experience in data center cooling, Kevin brings invaluable insights and thought leadership to the liquid cooling space.

The post The Real Challenges of Liquid Cooling Begin After Installation appeared first on Data Center POST.

TL;DR

• Cameras used as visual sensors can turn physical conditions into useful operational data, monitoring equipment between inspections and reducing routine manual checks.
• Connecting visual and thermal data with BMS and DCIM platforms gives engineers context to assess alerts sooner, provided hardware and integration suit the task.
• Start with a familiar monitoring task and test whether it reduces verification time and unnecessary visits before extending deployment.

# # #

By Peter Dempsey, EMEA Key Account Manager & Data Centre Lead, Axis Communications

Data center engineers spend part of their working day collecting information that equipment could, in some cases, report for itself. Reading a gauge or checking equipment with a handheld thermal camera remains valuable work, but repeating those tasks across a large facility takes time. Much of that effort is necessarily spent confirming that conditions remain normal – work that increasingly lends itself to automation.

Used as visual sensors, cameras can assist with much of this routine observation. They can provide a continuous view of physical conditions, while suitable analytics turn a gauge reading or thermal change into data that operational systems can use. In doing so, they connect what is happening at the equipment with the platforms responsible for monitoring it, including conditions that might otherwise require someone to visit and verify.

The opportunity is therefore not simply to add more monitoring, but to make physical evidence useful within the systems operators already trust. That means deciding where visual sensors can add value, what information they should provide and how that data will integrate with the wider monitoring strategy. 

Make the information useful where it arrives

For most operators, that means starting with the building management system (BMS) and data center infrastructure management (DCIM) platform. These are already central sources of operational oversight. Visual and thermal information adds value when it appears within those workflows; if it remains in a separate interface, engineers still have to find it and manually match it to the equipment concerned.

Consider a temperature alert associated with a particular cabinet. Linking it to the relevant thermal view allows the operator to remotely examine the area where the abnormal condition is concentrated. A suitable visible-light view may provide further context, helping them assess what needs investigating before they attend.

The connection should work in both directions. Status information from a rack or management platform can accompany the view of that asset, so the engineer can relate what they see to what the equipment is reporting. 

BMS and DCIM platforms retain their management roles, but benefit from additional information available to support decisions. That information can also generate, or be attached to, a work order or help a remote engineer advise colleagues on site. 

The practical test is to determine whether the addition of visual data actually reduces the work needed to understand the event. If staff still have to search several systems before they can assess it, the integration has left an important part of the job unfinished.

Monitor what changes between inspections

A scheduled inspection confirms conditions only at the time of the visit. It cannot account for a battery string that starts running hot shortly afterwards, or airflow that deteriorates following maintenance. In this case, detection depends on a secondary monitoring system or, often, on the next person arriving.

Increasing round frequency reduces that interval, at the cost of additional staff time. For teams responsible for expanding facilities, this becomes difficult to sustain. Operators are now in the position of identifying which checks would benefit from continuous observation and which still require an engineer’s physical presence.

Fixed thermal monitoring can complement periodic surveys by watching selected equipment between visits. Where the device and application support it, a temperature exception can trigger an alert accompanied by a view of the affected area. The engineer can then assess the incident’s location and extent when deciding how urgently to respond.

Advancements in monitoring do not replace expertise; a heat spike still needs interpretation. It may indicate a failing component, obstructed airflow or a condition that has already corrected itself. While thermal and visual information can help quickly narrow the investigation, neither will explain every fault.

Gauge monitoring offers another practical application. If the current process requires someone to read a dial and enter the result into a report, a suitable camera and reading application can instead supply that value directly to the management platform. The facility gains an auditable record that can be compared with defined limits, without waiting for the next round.

Choose hardware that supports the task

The monitoring task should determine the hardware, not the other way around. Reading a gauge requires a camera and lens that capture enough detail at the installed distance, with a clear viewing angle and lighting that avoids glare across the dial. Direct temperature monitoring requires a thermal device designed to provide temperature measurements across the requisite range, with sufficient accuracy to meet defined thresholds.

Processing capability also needs to match the application. If analytics are to run on the device, it must support that software and have the resources to perform the required task. The task itself must be clearly defined before procurement. Monitoring a thermal area against agreed limits, for example, gives operators something they can test against actual equipment conditions.

Specify the connection as well as the device

Before selecting equipment, operators must also establish how its output will reach the receiving platform. The system needs to recognize what a reading represents and which asset it belongs to. Putting devices on the same network does not, by itself, make their information useful to one another.

Protocols such as MQTT, Modbus TCP and OPC UA can support data exchange between visual sensors and management platforms, so operators should confirm which interfaces are supported and whether integration software is needed between them. But a protocol listed on a specification sheet is only part of the answer. An open platform gives operators greater freedom to build the integrations they need without the hardware itself becoming the limiting factor. 

The integration must also be secured, monitored and maintained throughout its operational life. Clear failure alerts are essential: an automated check should never give operators misplaced confidence in a reading that is no longer accurate.

Start with a task the team knows well

A useful first deployment should address a recurring task that the operations team can describe in detail. Beginning with selected thermal checks or manually recorded gauges offers a manageable scope and existing known routines against which to compare performance. The trial should cover the equipment, but also what happens after detection, including how the engineer receives the information and responds.

The team can then assess whether the change reduces verification time or unnecessary visits, as well as whether it detects the intended conditions reliably. They should also check what happens when a view is obstructed or a reading cannot be obtained. Those situations establish where manual inspection must remain and how staff will know it is needed.

Before extending the deployment, operators should ask engineers what work the system has removed and what work it has added. Their experience, alongside the trial results, should guide the next investment. If the result is simply another screen to monitor, the integration has not yet delivered its intended value.

# # #

About the Author

Peter Dempsey has spent 25+ years in the security sector undertaking various roles that have seen him take responsibility for key elements of mission critical, high-profile projects across a number of different vertical markets. For the last two years Peter has focused his attention on how security technologies best support a business’s security strategy, whilst driving the adoption and elevating the importance of cybersecurity and emerging technologies for physical security practitioners in the Data Centre vertical. As the Industry Liaison Manager for Axis Communications, Peter has been instrumental in driving the Axis strategy for Data Centres across Northern Europe and the wider EMEA region. Peter’s passion for Data Centres can be traced back to the huge success in Ireland with Hyper Scale and Multi-Tenant Data Centre (MTDC) deployments. The key to this success was Peter’s commitment to deliver a reliable sustainable solution that will mitigate downtime.

The post Turn Visual Monitoring Into Operational Intelligence in the Data Center appeared first on Data Center POST.

TL;DR

• Performance-Driven Adoption: Growth is largely fueled by AI/ML workloads, big data analytics, and real-time transactions that require maximum compute performance without virtualization overhead.
• Security & Hybrid Integration: Strict privacy regulations in BFSI and healthcare favor single-tenant isolation, while hybrid/multi-cloud models let enterprises pair bare metal seamlessly with virtual public cloud environments.
• Global On-Demand Infrastructure: MOD Mission Critical’s MOD PaaSPort™ delivers dynamically programmable physical hardware and colocation from core to edge, offering low-latency connectivity across an ecosystem of cloud and network providers.

# # #

Written by iMiller Public Relations on Behalf of MOD Mission Critical.

Demand for bare metal infrastructure is experiencing significant expansion driven by the rigorous requirements of modern workloads, including artificial intelligence, machine learning, big data analytics, and real-time financial transactions. These data-intensive applications require consistent high performance, predictable control, and dedicated resources without the performance overhead inherent in virtualization.

On top of that, strict compliance and data privacy regulations in sectors like healthcare and banking, financial services, and insurance are accelerating the preference for isolated, single-tenant environments. Combined with the rise of hybrid and multi-cloud strategies, organizations are actively combining bare metal hardware with public cloud environments to optimize performance according to specific workload demands.

Financial projections underscore a steep growth curve across the global bare metal cloud services market. Fortune Business Insights estimates market value at over $14 billion in 2025, forecasting it to surpass $78 billion by 2034 at a compound annual growth rate (CAGR) of roughly 21%. Straits Research reflects a similar trend, projecting growth from just over $12 billion in 2025 to nearly $62 billion by 2034, with a projected 21% CAGR between 2026 and 2034. Although research methodologies differ, both analyses indicate that dedicated physical infrastructure is becoming a core strategic component for mainstream enterprise workloads rather than remaining confined to legacy or niche use cases.

MOD Mission Critical addresses this market demand through its MOD PaaSPort™ platform, delivering dynamically programmable, on-demand physical infrastructure and colocation from core to edge across global markets. MOD’s footprint spans major data centers to supply direct or ultra-low-latency connectivity to cloud, neocloud, and network providers. By integrating bare metal compute with managed colocation, networking, and cloud access, the platform enables businesses to flexibly source, deploy, and scale dedicated infrastructure suited for AI processing, regulatory compliance, and high-control environment needs.

To continue reading, please click here.

The post Why Bare Metal Infrastructure Is Gaining Ground appeared first on Data Center POST.

TL;DR

• Since 2011, the firm has supported companies across data centers, networks, subsea systems, infrastructure investment, technology, and AI.
• iMPR’s services extend beyond traditional PR to include strategic advisory, crisis communications, and community engagement.
• The firm’s Groundswell™ methodology helps organizations build transparency, alignment and trust with communities impacted by digital infrastructure development.
• Founder and CEO Ilissa Miller also leads the OIX Association Digital Infrastructure Framework Committee.

# # #

As artificial intelligence drives greater demand for data centers, power and connectivity, communications, and community engagement are becoming increasingly important to digital infrastructure development. iMiller Public Relations (iMPR) is marking its 15th anniversary as the industry enters its next phase of growth.

Founded in 2011 by Ilissa Miller, iMPR has grown alongside the digital infrastructure sector, supporting data center developers, network operators, subsea cable systems, infrastructure investors, technology providers, and AI innovators. Over the past 15 years, the firm has worked with companies representing hundreds of billions of dollars in infrastructure investment while generating thousands of media placements and thought-leadership contributions.

Beyond Traditional Public Relations

As the industry has evolved, iMPR’s capabilities have expanded beyond traditional media relations. The firm provides strategic advisory, message platform development, crisis communications, media and analyst relations, event and trade show marketing, and community engagement programs.

“At iMPR, our success has always been rooted in industry knowledge, relationships, and the ability to see where the digital infrastructure sector is heading,” said Ilissa Miller, Founder and CEO of iMiller Public Relations. “As AI, power, connectivity, and large-scale data center development reshape the global economy, communications, messaging accuracy, and community engagement are more important than ever. We are proud to celebrate 15 years supporting the companies building the infrastructure that powers the digital world.”

Community engagement has become a growing part of that work as data center developers navigate increased attention surrounding land use, power, resources, and local impact. iMPR’s Groundswell™ methodology helps organizations build transparency, alignment, and trust with communities impacted by digital infrastructure development.

Supporting Responsible Infrastructure Development

That focus extends beyond individual projects through Miller’s work with OIX Association, where she leads the Digital Infrastructure Framework Committee (DIFC). The committee is focused on advancing responsible infrastructure development by providing community planners with a framework to integrate digital infrastructure into their master plans.

Behind iMPR’s work is a team of more than 20 communications, marketing and digital specialists supporting client strategy, media relations, events, and partnerships. The leadership team includes Danielle DeMicco DiGisueppe and Jaclyn Litwa, Directors of Accounts; Brian Meeley, Director of Media & Analyst Relations; Kwan Fai Cheng, Director of Digital Operations; and Dillan Conn, Director of Strategic Initiatives and Business Platforms.

Over the past 15 years, iMPR has also built relationships with industry organizations, conferences, and associations across the digital infrastructure ecosystem. These connections help the firm support clients as they engage with industry stakeholders, policymakers, and local communities.

As AI, energy constraints, and community expectations continue to reshape digital infrastructure development, iMPR is focused on helping organizations communicate their vision, build trust with stakeholders, and navigate the evolving digital infrastructure landscape.

Learn more about iMiller Public Relations and its strategic communications, public relations and community engagement services by visiting www.imillerpr.com.

The post iMiller Public Relations Marks 15 Years Supporting Digital Infrastructure appeared first on Data Center POST.

TL;DR:

• AI workloads are pushing rack power densities from roughly 12 kW in 2024 to 26–27 kW today, with some AI racks already running 50-100+ kW.
• Unlike steady-state IT loads, AI training and inference create rapid, repeated power swings – not just backup events.
• Traditional backup batteries and capacitor-based systems each address different requirements: one prioritizes runtime, while the other prioritizes rapid response.
• Rack-level energy storage is emerging as its own design category, distinct from centralized UPS systems.
• Battery chemistry choices at the rack are now a safety and reliability variable, not just a capacity spec.

# # #

For most of the last decade, data center power design centered on one question: How long can the UPS ride through an outage? That question hasn’t gone away, but now it has company. ZincFive^®, known for nickel-zinc (NiZn) battery technology in UPS systems, announced an expansion into in-rack energy storage with its new NiZn In-Rack Power Solutions, which includes dedicated Battery Backup Unit (BBU) and AI Dynamic Power Module (DPM) configurations built for AI’s faster, more dynamic power draw. The move points to a broader shift underway across the industry: Rack-level power is becoming its own design discipline, separate from the UPS layer.

Average rack densities have more than doubled in two years, climbing from around 12 kW in 2024 to roughly 26-27 kW in 2026, according to recent industry benchmarking. Some AI training racks already exceed 50 kW, and next-generation GPU architectures are being designed around 100 kW or more per rack. That’s the headline most operators already know.

What’s less understood is how that power behaves. AI training and inference don’t draw a steady current; they spike and drop in fast, repeated bursts as GPU clusters ramp up, pause, and resume. These transient loads can stress busbars, power supplies, and upstream infrastructure in ways a traditional steady-state design never had to account for.

Backup Runtime and Rapid Response Are Different Problems

This is where the engineering gets interesting. Traditional battery backup units are built to provide runtime, with enough stored energy to ride through a utility disturbance for seconds to minutes. Capacitor-based systems, by contrast, respond in milliseconds but hold very little energy, so they smooth short spikes without offering meaningful backup duration.

Most facilities have defaulted to choosing one or blending both, accepting the tradeoff. A capacitor bank might absorb a millisecond-scale current step from a GPU cluster starting up, but it can’t cover a multi-second grid disturbance. A traditional BBU can cover that disturbance but may not be designed to manage rapid, repeated load transitions.

“AI is changing not only how much power data centers require, but how that power is consumed,” said Tod Higinbotham, CEO of ZincFive. “As power densities increase and workloads become more dynamic, the industry needs new approaches that can deliver rapid power response without compromising safety, reliability, or flexibility.”

As racks get denser and workloads more dynamic, that tradeoff becomes harder to accept. Operators need infrastructure that can do both jobs without over-provisioning space or cooling capacity at the rack, which is already at a premium in AI deployments.

Why Battery Chemistry Matters at the Rack Level

As energy storage moves physically closer to compute, the chemistry inside that storage becomes a bigger factor in facility risk management. Lithium-ion remains dominant across the industry, but proximity to high-density, high-heat compute racks raises the stakes around thermal behavior.

Nickel-zinc is one proven chemistry gaining attention specifically for this reason, already validated at scale in UPS-level deployments before moving closer to the rack. It supports high-rate charge and discharge capabilities, which suit repeated transient events, while carrying no risk of thermal runaway. For rack-level deployments, where space constraints can limit fire suppression and containment options, that safety profile is becoming as relevant as energy density or cycle life.

What This Means for AI Infrastructure Planning

The practical takeaway for facility planners and OEMs is that rack-level power is no longer a single-spec decision. It requires evaluating backup runtime, transient response speed, thermal safety, and physical footprint together, not in isolation.

ZincFive is now developing dedicated in-rack solutions – BBU configurations for established backup applications and AI DPM configurations for transient and pulse power mitigation – rather than a single configuration meant to address every requirement. That approach reflects a broader industry shift: As AI reshapes how power is consumed, the tools used to manage it are having to specialize rather than generalize.

For an industry accustomed to sizing infrastructure around steady-state assumptions, that’s a meaningful adjustment, and one that’s likely to define rack design conversations as AI infrastructure continues to evolve.

To read the full press release, visit here.

The post Rethinking Rack-Level Power for AI Data Centers appeared first on Data Center POST.

TL;DR

• Data centers support users globally, but their impacts on land, electricity, water, tax base and environmental resources are experienced at the local level.
• Communities should incorporate digital infrastructure into comprehensive planning before individual development applications arrive.
• Responsible planning should account for data centers as well as fiber routes, network nodes, substations, transmission infrastructure and other systems supporting the digital economy.
• Effective governance requires early community engagement, transparency, regional planning and consideration of a project’s full lifecycle.
• The OIX Digital Infrastructure Framework is designed to help municipalities proactively evaluate digital infrastructure needs and development considerations.

# # #

By Ilissa Miller, CEO of iMiller Public Relations

We talk about “the cloud” as though it exists somewhere above us: intangible, limitless and disconnected from the physical world. Of course, it isn’t.

The cloud is buildings, servers, fiber, substations, cell towers, transmission infrastructure, cooling systems, land, power, water, and increasingly, communities trying to understand what all of this means for them. That reality was at the center of the Data Centers and AI panel I recently participated in at IGF-USA 2026 in Philadelphia.

Moderated by Daniel Castro, President of the Information Technology and Innovation Foundation (ITIF), the discussion included Tana Fitzpatrick of the University of Oklahoma, Chris Amato of the Lehigh Valley Planning Commission, Maya Schmidt of Third Way and me, Ilissa Miller, representing iMiller Public Relations and the OIX Digital Infrastructure Framework.

Daniel framed the issue well from the outset: data centers support users around the world, but their impacts: land, electricity, water, tax base and environmental considerations are experienced locally.

So the question isn’t simply whether we need data centers. The question is how we govern their development.

Plan Before the Application Arrives

My position is that communities shouldn’t be forced to figure this out one development application at a time. Digital infrastructure needs to become part of comprehensive community planning.

That means identifying appropriate land uses for different types of data centers, but it also means looking beyond data centers to fiber routes, network nodes, substations, transmission infrastructure and the other systems communities will need to participate in an increasingly digital economy.

Not every data center belongs everywhere. And not every community needs the same infrastructure. But communities should understand their needs before a developer arrives with a proposal.

That’s one of the reasons we developed the OIX Digital Infrastructure Framework: to give municipalities a way to proactively understand these issues rather than continually reacting to them.

Start With the Community

Tana Fitzpatrick added an incredibly important dimension to the conversation: tribal sovereignty.

Federally recognized tribal nations are governments, not simply another community stakeholder. As Tana explained, engagement shouldn’t begin by asking what land, water, energy or other resources a tribal nation has that a project might need.

It should begin by understanding what that nation and its community actually want. Her question was wonderfully simple: What does being a good neighbor look like? You can read more about her work at the University of Oklahoma Native Nations Center here.

Look at the Regional Picture

Chris Amato brought the planning perspective from Pennsylvania’s Lehigh Valley, where communities are already confronting multiple proposed developments.

His perspective reinforced something I believe strongly: site and setting matter.

Brownfields, existing industrial properties and adaptive reuse may offer very different opportunities than greenfield development. Communities also need to evaluate infrastructure capacity, workforce benefits, environmental resources and net fiscal impact — not simply the large investment number announced with a project. You can read more about the Lehigh Valley Planning Commission here.

Make the Benefits and Responsibilities Tangible

Maya Schmidt of Third Way focused on another important component: ensuring communities have the information and leverage necessary to negotiate effectively.

Transparency, community engagement, infrastructure costs, tax contributions, workforce development and community benefit agreements can all be part of that conversation. You can read Maya’s Third Way data center research here.

I would add one more element: the entire lifecycle. What happens if ownership changes? Do commitments made to the community transfer to the next owner? What happens when equipment becomes obsolete? What are the decommissioning responsibilities? What happens when better technology becomes available?

Responsible development can’t end when the permit is issued.

Social License Starts With Listening

Ultimately, perhaps the simplest takeaway from our discussion was this: Developers need to listen before they propose. And, they need to start that listening process much sooner than they think.

Go into the community. Meet people. Understand their concerns. Learn what they value. Then demonstrate how that input changed the project. As I said during the discussion, development needs to become of the community, for the community and by the community.

The data center industry has an opportunity right now to redefine how large-scale infrastructure development happens in America. Not by dismissing opposition, and certainly not by assuming every project belongs everywhere. But by planning earlier, listening better, being transparent and creating guardrails that communities and industry can understand together.

That isn’t simply better community engagement. It’s better governance.

Curious about what you missed? Not to worry, you can still view the live stream online here: isoc.live/21744.

Stay tuned for IGF-USA upcoming events: igfusa.us.

The post The Cloud Has a ZIP Code: Governing the Infrastructure Behind AI appeared first on Data Center POST.

TL;DR

• Surging AI Energy Demand: AI expansion is driving a massive surge in data center electricity consumption, with U.S. data center usage projected to reach between 9.5% and 15.3% of total domestic power by 2030 and global consumption expected to roughly double to 945 TWh
• Infrastructure Timing Mismatch: Traditional grid interconnection approaches are encountering severe bottlenecks because data centers can be developed in 2 to 3 years, whereas major utility power infrastructure takes considerably longer to plan, permit, and build
• Active Role for Battery Storage: Battery Energy Storage Systems (BESS) are evolving from passive emergency backups into core drivers of facility power architecture, working alongside microgrids and on-site generation (solar, fuel cells, natural gas) to manage peak demand and ensure mission-critical reliability
• Early Civil and Regulatory Integration: Deploying utility-scale storage requires early cross-functional alignment, as project success is frequently dictated by physical site conditions, civil balance-of-plant work (such as drainage and stormwater management), fire safety, and local authority reviews

# # #

By William J. Pirrone, P.E., Founding Principal, Rubicon Professional Services

Most people are aware that AI is changing and expanding the data center industry. What gets less public attention is what this growth means for the infrastructure needed to power these data centers.

I have spent more than two decades working on critical facilities and energy infrastructure, and one of the biggest changes I have seen is the amount of power developers now need and how quickly they need it. The traditional approach of designing a facility and securing a utility interconnection is becoming increasingly difficult, putting sustainable infrastructure planning at the center of nearly every new project.

According to the U.S. Department of Energy and Lawrence Berkeley National Laboratory, data centers accounted for about 4.4% of U.S. electricity consumption in 2023. Their 2024 study projected that share could reach between 6.7% and 12% by 2028. A recent Berkeley Lab update estimates that data centers could account for 11.8% of U.S. electricity use by 2030, with scenarios ranging from 9.5% to 15.3%. The two studies indicate that these issues are also playing out globally. Case in point: The International Energy Agency projects that electricity consumption by data centers will roughly double by 2030, to about 945 terawatt-hours. AI is the biggest driver of that increase.

This research doesn’t represent an abstract energy-policy discussion; it dictates where and how data centers can be built, how quickly capacity can come online and, ultimately, whether a project is economically viable. Increasingly, developers are turning to energy storage technology deployed alongside the grid rather than as a backup to address this new industry dynamic.

Battery Storage Is Moving To The Forefront

For years, battery energy storage in critical facilities was primarily associated with backup power and UPS systems, but that’s no longer the case. It is becoming a main driver of power decisions.

Today’s battery energy storage systems (BESS) play a much more active role in the facility’s power architecture. Properly designed, BESS help manage peak demand, provide rapid response when grid conditions change, and work alongside other sources of generation.

To be properly optimized, a data center’s power needs to run through integrated energy management systems that combine grid power, battery storage, and on-site generation, including solar, fuel cells, and natural gas. Today’s objective is to create a power system that is reliable enough for mission-critical computing but flexible so that the facility isn’t entirely dependent on a single power source.

A study commissioned by California’s three major investor-owned utilities found that a modeled solar facility paired with four hours of battery storage could achieve a 99.8% effective load-carrying capability. This study demonstrated how dramatically storage can improve the reliability value of intermittent generation.

Recent improvements in battery technology advancements have made systems like this possible, but they have also made them more complex. I saw that firsthand recently on a 200-megawatt, multi-hour storage project at an existing power generation site. The system wasn’t a simple containerized battery; it was closer to a distributed industrial process plant, with its own water treatment, electrolyte handling, and air management systems layered on top of the electrical framework. Evaluating it meant pricing out civil work, mechanical systems, and electrical balance-of-plant as separate, substantial line items, rather than add-ons to a battery purchase.

However, every modern data center should not simply install solar panels and batteries as this one did. Site conditions, load profiles, utility requirements, and economics vary considerably. But this 200-megawatt, multi-hour storage demonstrates that storage has moved well beyond its traditional role as an emergency backup resource.

Leveraging Batteries: Experience Installation Matters

A utility-scale battery installation is not another piece of IT equipment that can simply be added to a data center. Large battery systems call for careful attention to fire protection, electrical safety, commissioning procedures, and emergency response. They also require close coordination with utilities, fire officials, inspectors, and other authorities, and some communities may be dealing with the review of a large-scale battery installation for the first time.

On one recent project, the single biggest cost and schedule driver wasn’t the battery technology at all; it was a pair of stormwater retention ponds already on the site. Adjusting to the new installation required partial infill, regrading, and a redesigned drainage system, all before a single battery enclosure could be installed. It’s a good reminder that the constraint that decides whether a site works often isn’t the equipment—it’s what’s already in the ground.

AI Is Forcing A Power Strategy Change

A few years ago, power infrastructure was often treated as something that followed the data center design. Today, designers need to consider power much earlier. Power availability, interconnection timing, on-site generation, and energy storage will determine a facility’s success.

The International Energy Agency has pointed out the basic timing problem: a data center can potentially be developed in two or three years, while major energy infrastructure can take considerably longer to plan, permit, and build.

The Future Should Not Contain Mismatched Power

The data centers that support the next generation of AI will require enormous amounts of reliable electricity. The industry cannot assume the grid alone will always provide that power on the timetable developers need. That is why battery storage, microgrids, and behind-the-meter generation are becoming so important. The facilities that succeed will be those where engineering, construction, utilities, local authorities, and the providers of these capabilities are brought together early enough to turn a power strategy into something that thrives in the field.

# # #

About the Author

William Pirrone is founding principal of Rubicon Professional Services and Rubicon Technical Services. He brings more than 35 years of engineering design and construction, operations, and maintenance experience to the data center, telecommunications, and critical power industries, where he established a significant presence. William’s technical expertise and understanding of an owner’s perspective have been instrumental in the rapid success of both RPS and RTS on a national level.

As a founder of the Rubicon companies, William has developed a unique set of service offerings specific to mission critical environments, designed to fill the void left by traditional construction companies and methods. These services include technical program management; construction management; equipment integration; engineering, procurement, and construction (EPC); commissioning; and energy management services.

Prior to establishing the Rubicon companies, William was the director of global engineering at JP Morgan Chase, where he led the design and integration of all data, trading, and critical operating centers through multiple billion-dollar mergers and acquisitions globally. That charge carried across 60 countries and more than 70 million square feet of real estate. While at JP Morgan Chase, William also provided direct support to the real estate division by reviewing and negotiating leases for key operations centers, developing tactical and strategic facility restack/disposition planning, and line of business relocation strategies at the corporate level. He also partnered with the energy procurement team to secure substantial cost savings and pricing stability in the deregulated energy markets across the United States. William departed JP Morgan Chase to launch Rubicon. He holds a B.S. degree in electrical engineering from Pratt Institute, is a licensed Professional Engineer, and holds multiple general contractor licenses.

The post Powering the AI Economy: What the Future Holds for Data Center Energy Infrastructure appeared first on Data Center POST.

TL;DR

• The Internet Governance Forum (IGF) provides a multistakeholder platform for government, industry, the technical community, academia and civil society to discuss public policy issues surrounding the Internet.
• The United Nations General Assembly made the IGF a permanent UN forum in December 2025, shifting the conversation toward what the forum should become next.
• Internet governance now extends beyond domain names and IP addresses to artificial intelligence, cybersecurity, privacy, digital inclusion and the physical infrastructure supporting the digital economy.
• Industry has a responsibility to bring technical knowledge and operational experience to governance discussions while listening to policymakers, communities and other stakeholders.
• The OIX Digital Infrastructure Framework provides communities and developers with a common foundation for evaluating digital infrastructure and understanding responsible development.

# # #

By Ilissa Miller, CEO of iMiller Public Relations

I recently had the opportunity to participate in IGF-USA 2026 at the Museum of the American Revolution in Philadelphia. I was invited to share the OIX Digital Infrastructure Framework, an initiative I’ve been leading to help communities better understand, plan for and govern the digital infrastructure increasingly becoming part of their physical landscape.

I expected a thoughtful discussion about Internet governance. What I didn’t fully expect was how directly the conversations would intersect with many of the challenges our industry is confronting today.

First, What is the IGF?

The Internet Governance Forum, or IGF, was established by the United Nations as a multistakeholder forum for discussing public policy issues surrounding the Internet. It brings together government, industry, the technical community, academia and civil society — not to negotiate regulation, but to develop greater understanding of complicated issues and inform those who ultimately make policy and business decisions.

The Internet wasn’t built by any one government, company or organization. It evolved through collaboration among technologists, network operators, businesses, policymakers, researchers and users. It makes sense that its governance should benefit from that same diversity of expertise.

And the timing couldn’t be more important.

In December 2025, the United Nations General Assembly made the IGF a permanent UN forum. The conversation has therefore shifted from whether IGF will continue to what it should become.

As one participant memorably put it during IGF-USA, the IGF has finally “caught the bus.” After 20 years of working toward permanence, the question becomes: Now what?

Relaunching IGF-USA

That question was particularly relevant because this year’s gathering also marked the relaunch of IGF-USA.

Led by Ron da Silva, Chair of the IGF-USA 2026 Steering Committee, the Philadelphia event brought together an impressive cross-section of stakeholders both in person and virtually.

The multistakeholder participation was one of the things that impressed me most. These weren’t simply people talking about technology. They represented organizations and communities actively dealing with how technology affects policy, infrastructure, economic development, security and society.

The event was supported by Amazon, ARIN, ICANN, Identity Digital, the Internet Society Foundation and Verisign — organizations representing very different pieces of the Internet ecosystem, and it was live streamed here: isoc.live/21744

From Technology to Governance

The Internet governance conversation has also changed dramatically.

Early IGF discussions focused heavily on domain names, IP addresses and the underlying architecture of the Internet. Today, the agenda includes artificial intelligence, cybersecurity, privacy, digital inclusion and increasingly the physical infrastructure supporting our digital economy.

AI makes this conversation even more urgent.

Technology is advancing faster than many traditional policymaking processes can reasonably absorb. Yet governments are increasingly being asked to make decisions about technologies, infrastructure and markets they may not fully understand.

That’s why the multistakeholder model matters.

Industry cannot simply show up to tell policymakers what it wants. We need to bring technical knowledge, operational experience and an understanding of real-world implementation — and we also need to be prepared to listen, adapt and be challenged. That was a recurring theme throughout the day.

We Have a Responsibility to Help Govern What We Built

That was probably my biggest takeaway from IGF-USA.

Those of us who have spent our careers building the Internet and digital infrastructure have a responsibility to help establish the guardrails for what comes next.

We understand how these systems work. We understand how infrastructure gets deployed. We understand where networks, data centers, energy systems and increasingly AI intersect.

We therefore cannot wait for regulation to solve every problem.

We need to self-govern where appropriate. We need principles. We need transparency. We need accountability. And we need mechanisms for industry, government, communities and other stakeholders to understand one another before disagreements turn into policy battles.

That is also why I believe the OIX Digital Infrastructure Framework belongs in this conversation. The Framework is intended to create a common foundation for communities and developers to evaluate digital infrastructure, giving both sides better questions to ask and a better understanding of what responsible development should look like.

The IGF does not need to become another regulatory body. Its value may be precisely the opposite: creating a place where complicated issues can be understood before decisions are made.

With AI accelerating the next evolution of the Internet, we need those conversations more than ever.

Curious about what you missed? Not to worry, you can still view the live stream online here: isoc.live/21744.

Stay tuned for IGF-USA upcoming events: igfusa.us.

The post What Is IGF and Why Internet Governance Matters More Than Ever appeared first on Data Center POST.

TL;DR

• Communities should include digital infrastructure in long-term land-use and development plans before individual projects are proposed.
• Early engagement allows developers to understand local priorities and incorporate community feedback into project designs.
• Concerns about data centers increasingly include broader anxieties surrounding artificial intelligence, resource use, and community impact.
• The OIX Digital Infrastructure Framework helps municipalities identify the questions they should consider when evaluating digital infrastructure.
• Responsible development begins with determining whether a project is appropriate for a particular community and location.

# # #

Data center development is accelerating as artificial intelligence creates greater demand for computing capacity, power, and supporting infrastructure. At the same time, many municipalities are being asked to evaluate projects that were never anticipated in their zoning codes or long-range development plans.

A recent episode of the Data Center Frontier Show explored how communities can prepare for this growth before a specific project arrives. Host Matt Vincent spoke with Ilissa Miller, Founder and CEO of iMiller Public Relations and a board member of the Open Infrastructure Exchange (OIX), about proactive planning, community engagement and the OIX Digital Infrastructure Framework.

Communities regularly use comprehensive plans to determine where residential, commercial, and industrial development should occur. Digital infrastructure, however, is often missing from those discussions. Even land already designated for industrial development may not explicitly allow data centers, creating uncertainty for residents, developers, and public officials when a project is proposed.

The OIX Digital Infrastructure Framework is intended to help municipalities consider these issues earlier. Rather than advocating for the approval of specific projects, the framework helps local leaders evaluate land use, infrastructure capacity, and community needs while identifying which questions fall under municipal authority and which require coordination with utilities or other government entities. Proactive planning also allows communities to distinguish among different forms of digital infrastructure. A municipality may determine that it can support a large AI campus, while another may be better suited for colocation, edge infrastructure, or smaller facilities serving local businesses and institutions.

Community opposition to data centers increasingly extends beyond concerns about power, water, noise, construction, and taxes. Data centers have also become highly visible symbols of artificial intelligence, causing broader questions about automation, surveillance, and AI governance to enter local land-use debates.

Those concerns should be acknowledged, but communities must also understand the specific decision before them. Questions about national or statewide AI policy are different from determining whether a particular property is appropriate for digital infrastructure. Clearly explaining the proposed facility, its potential impacts, and the stages of the approval process can help residents better understand what is and is not being decided.

For developers, meaningful community engagement should begin before a conditional-use application is submitted or a project design is finalized. This process includes studying the community, understanding the history of the proposed site and identifying the priorities of residents, businesses and local officials.

Listening sessions, open houses, business forums, and community webinars can help developers gather that information. However, engagement is most effective when community feedback can still influence the project. Developers should explain what they heard, identify what changed, and provide increasingly detailed information as plans progress.

Community benefits should also reflect local priorities rather than follow a standard formula. Depending on the community, residents may prioritize workforce development, education, transportation improvements, tax revenue, environmental protections, or preservation of local land.

Community acceptance is not solely a communications challenge. Developers must also determine whether a project belongs on the proposed property. Some locations may be unsuitable because they require significant conversion of agricultural or other valued land, particularly when existing industrial properties or brownfield sites are available.

This means responsible development may sometimes involve deciding not to move forward. Evaluating land use, infrastructure availability, and community needs before acquiring or pursuing a site can reduce conflict and help ensure that projects align with local planning goals.

To hear the full conversation with Ilissa Miller on preparing communities for AI data center development, listen to the Data Center Frontier Show podcast episode here.

The post Planning for AI Infrastructure Before Data Center Projects Arrive appeared first on Data Center POST.

TL;DR

• 365 Data Centers volunteered with Boca Helping Hands, a Palm Beach County nonprofit providing food, medical, and financial assistance, as part of a community give-back day near its Boca Raton data center.
• The event highlights 365’s continued investment in the local communities surrounding its 16 U.S. data center markets, beyond its role as an infrastructure provider.

# # #

365 Data Centers, a leading provider of high-density colocation, connectivity, and cloud services, recently spent a day giving back to the community surrounding its Boca Raton data center. Team members volunteered with Boca Helping Hands, a nonprofit organization dedicated to fighting hunger and poverty throughout Palm Beach County, Florida.

Boca Helping Hands provides food assistance, medical care, and workforce development programs to help local individuals and families build long-term stability. In 2025 alone, the organization logged nearly 58,000 volunteer hours and distributed more than 3.5 million meals to residents in need.

For 365 Data Centers, which operates a facility within the Boca Raton area, the volunteer effort underscores a broader philosophy that extends beyond powering enterprise infrastructure. The company considers serving the communities where it operates a core part of its identity as a regional and national infrastructure provider.

“Great day spent volunteering with the team at Boca Helping Hands,” said 365 CSO Charlie Dankner following the event. “This organization is a well-oiled efficiency machine that provided almost 3.5 million total meals in 2025 to families in need. That included over 288,000 weekend meals for kids while generating 54 new jobs via their workforce development programs. Proud to give back to the communities we serve.”

The Boca Raton volunteer day adds to 365 Data Centers’ growing record of local community involvement across its 16 U.S. data center markets, reinforcing the company’s presence not just as an infrastructure partner, but as an engaged neighbor. For more information on 365 Data Centers’ locations and service offerings, visit 365datacenters.com.

The post 365 Data Centers Gives Back to the Boca Raton Community with Boca Helping Hands appeared first on Data Center POST.

TL;DR

• Securing AI workloads requires layered protection that covers the infrastructure, connections, data, models and the ongoing activities that support AI operations.
• AI workloads rely on specialized computing infrastructure, large datasets and interconnected systems, with demand for AI-related data center capacity growing rapidly.
• Stronger security is important because AI workloads concentrate sensitive data, valuable computing resources and interconnected systems within data center environments.
• A layered strategy combines physical security, network and access controls, data and model protection, data orchestration safeguards and continuous monitoring.

# # #

AI workloads are changing how data centers operate and how teams secure them. As these environments expand, security teams need to protect the infrastructure, data, models and connections that support operations. A layered security strategy gives data center teams a practical way to protect these workloads throughout their life cycle.

Understanding AI Workloads in Data Centers

AI workloads refer to the computing processes that support the development and operation of artificial intelligence systems. These activities include model training, fine-tuning, inference and data processing. They often rely on GPUs and other specialized accelerators that work across interconnected servers and storage systems.

The scale of this infrastructure is expanding rapidly. Global data center capacity demand will grow by three and a half times between 2025 and 2030 as AI workloads expand. The volume alone will require an estimated 156 gigawatts of capacity by 2030, representing over four times the projected level for 2025. 

AI environments also involve extensive data flows between compute resources, storage platforms, applications and external services. Training datasets need to reach processing systems efficiently, while completed models require secure access to applications and users. These requirements create an environment where performance, availability and security must work together across the data center.

Why AI Workloads Need Stronger Data Center Security

AI is changing how businesses operate across knowledge-based industries, and financial markets are reflecting concerns about the scale of that change. An analysis reported a sharp selloff in 2026 across software, data providers and other sectors. Price declines emerged amid growing concerns about AI-driven disruption, including advances in coding and business process agents. 

The market reaction illustrates the uncertainty surrounding AI-driven changes to established business models. Security risks add another layer of concern as organizations increasingly depend on AI workloads to support critical business operations. These workloads concentrate valuable data and intellectual property in highly connected environments, raising security considerations for the data centers that host them. 

• Sensitive data exposure: AI workloads may process proprietary business information, customer records, research, financial information and other sensitive datasets. Strong access controls and encryption help keep this information within authorized environments.
• High-value computing infrastructure: GPUs, accelerators, servers, storage systems and networking equipment represent significant computing resources. Unauthorized access or malicious activity affecting these components could disrupt AI operations and expose valuable systems.
• Expanded attack surface: AI workloads connect models, datasets, APIs, applications, storage platforms, networks and orchestration tools. Each connection introduces security requirements that need to remain aligned across the environment.
• Model and data integrity: Training data and model files directly influence AI outputs. Security controls need to protect these assets from unauthorized changes, manipulation or tampering throughout the AI pipeline.
• Operational and financial impact: Disruptions to AI training and inference can consume valuable computing resources, delay projects and affect applications that depend on AI services. Security measures help maintain reliable access to the infrastructure supporting these workloads.

How to Secure AI Workloads in Data Centers

Securing AI workloads requires several protective layers across the data center environment. Each addresses a different part of the workload life cycle and helps teams maintain reliable, controlled operations. A layered strategy also gives security teams multiple opportunities to detect and contain threats as AI environments grow.

Secure the Physical Infrastructure

The scale of AI infrastructure makes physical protection especially relevant. Meta operates 15 data center campuses across 21 U.S. states, with a combined capacity of 12,690 megawatts. The company is also expanding this infrastructure to support its growing AI operations, illustrating how AI computing can span large data center environments.

As AI infrastructure continues to expand across data center campuses, physical access restrictions become increasingly important. Teams should secure servers, GPU systems, storage equipment and networking hardware through controlled entry points, authentication systems, surveillance and access logging. 

Administrative access requires similar attention. Strong identity controls help ensure that only authorized personnel can manage critical infrastructure.

Security teams also need to maintain the underlying technology. Regular patching for servers, operating systems, firmware, GPUs, and related components closes known vulnerabilities and supports a more resilient environment. Hardened configurations and secure boot mechanisms can provide additional protection where the hardware supports them.

Protect Networks and Control Workload Access

AI workloads depend on fast connections between computing resources, storage, applications and users. Network security must protect these connections while meeting the performance requirements of AI operations.

Network segmentation can separate AI environments from unrelated systems and limit the movement of unauthorized activity. Least-privilege access also helps ensure that users, applications, services and machine identities receive only the permissions required for their roles.

For data moving between systems, encryption helps secure these transfers. Network monitoring also gives security teams visibility into activity across compute nodes, storage platforms, applications and external services. Restricting unnecessary inbound and outbound connections further reduces opportunities for unauthorized access.

Zero-trust principles provide another layer by requiring authentication and authorization for access based on established policies. These controls remain useful as AI workloads span multiple systems, environments and administrative teams.

Secure AI Data, Models and Data Orchestration

AI security also depends on protecting the information that feeds models and the assets that AI systems generate. Training, validation and inference data may pass through several systems before reaching their destination, creating a need for consistent safeguards throughout the pipeline.

Encryption helps protect sensitive information at rest and in transit. Access controls can restrict datasets according to user roles, workload requirements and data sensitivity. Data lineage also gives teams visibility into where information originated and how it moves through AI workflows.

Apart from storage, data orchestration helps manage how information moves through the systems that support AI workloads, including its transformation and preparation. Secure orchestration requires appropriate permissions, reliable logging and controls that help verify data as it moves between systems.

Monitor Workloads and Respond to Threats

AI workloads need continuous monitoring because their activity can change as models, applications, users and data pipelines evolve. Security teams should establish visibility across multiple areas of the system:

• Logging and audit trails provide records that help teams investigate unusual events.
• Monitoring GPU utilization, data transfers, authentication activity and workload behavior reveals patterns that warrant further review.
• Automating alerts helps security personnel identify potential threats quickly. 

Response procedures should define how teams isolate affected systems, investigate incidents, restore services and review the controls involved.

Build Security Into Every Layer of AI Infrastructure

AI workloads are becoming a major driver of data center capacity, creating a growing need for safeguards across the infrastructure that supports them. Organizations can strengthen AI workload security by establishing controls at each layer and maintaining visibility as systems evolve. Protecting the infrastructure, connections, data, models and workload activity helps create a secure foundation for continued AI deployment.

# # #

About the Author

April Miller is a Technology writer with more than 5 years of experience writing on AI applications that help SMBs compete and Enterprises strategize. You can connect with her on LinkedIn.

The post How to Secure AI Workloads in Data Centers appeared first on Data Center POST.

TL;DR

• As enterprises shift to hybrid and multi-cloud architectures, a data center’s connectivity options matter as much as its physical location.
• 365 Data Centers’ expanded partnership with Megaport adds nine new Points-of-Presence, bringing its connectivity footprint to 15 U.S. facilities.
• The move reflects a broader industry shift: colocation providers are increasingly judged on their connectivity ecosystem, not just power and space.

# # #

For years, choosing a colocation facility largely came down to three questions: How much power is available, how reliable is the site, and how close is it to my end users. Today, a fourth question carries just as much weight: How easily can I connect from this facility to the cloud platforms, networks, and digital services my business actually runs on?

That shift is playing out visibly in the market. 365 Data Centers recently expanded its partnership with Megaport, a global Network-as-a-Service (NaaS) provider, adding nine new Points-of-Presence across facilities in Fort Lauderdale, Detroit, Chicago, Commack, Philadelphia, Rancho Cordova, Richardson, Smyrna, and Herndon. The expansion brings Megaport’s footprint across 365’s platform to 15 facilities nationwide – a scale-up worth examining for what it signals about enterprise infrastructure decisions more broadly.

Why On-Demand Connectivity Matters More Than Ever

Enterprises rarely run on a single cloud anymore. A typical mid-market or enterprise IT environment might touch AWS for compute, Azure for productivity tools, and a specialized platform like Oracle Cloud or IBM Cloud for a legacy workload that hasn’t migrated yet. Managing separate physical connections to each of these providers used to mean long procurement cycles and rigid, hard-to-change infrastructure.

Software-defined connectivity platforms change that equation. Instead of provisioning a dedicated line to each cloud provider individually, customers can provision, scale, or retire connections on demand through a single interface. That flexibility matters most when business needs shift quickly, such as for a merger, a new regional office, or a sudden capacity spike shouldn’t require months of lead time just to get connected.

Edge Markets Are Where This Plays Out

It’s worth noting where these new connections are landing. Markets like Commack, New York, and Smyrna, Georgia, aren’t traditional hyperscale hubs; they’re regional edge markets, closer to where mid-sized businesses and regional offices actually operate. As AI workloads and latency-sensitive applications push more compute closer to end users, connectivity in these secondary markets becomes just as important as in major metros.

This is a trend worth watching across the industry: Colocation providers are extending carrier-neutral, multi-cloud connectivity into markets that previously lacked it, narrowing the gap between edge and core infrastructure.

What This Means for IT Decision-Makers

For organizations evaluating colocation partners, a few practical takeaways emerge from this kind of expansion:

• Ask about the connectivity ecosystem, not just the facility specs. A data center with excellent power density but limited network options can still create bottlenecks.
• Favor providers offering on-demand, software-defined connections over fixed, long-term circuits, especially if your cloud strategy is still evolving.
• Consider regional and edge locations carefully. As more colocation providers build out connectivity in secondary markets, businesses gain more flexibility in balancing latency, cost, and resilience.

The Bigger Picture

Partnerships like this one reflect a maturing colocation market, where facilities are increasingly valued as connectivity hubs rather than just physical real estate. As hybrid IT becomes the default rather than the exception, the providers that make multi-cloud connectivity simple – rather than something customers have to engineer themselves – are the ones best positioned to support enterprise growth in the years ahead.

Click here to read the full announcement.

The post How 365 Data Centers and Megaport Are Redefining What “Connected” Means for Colocation appeared first on Data Center POST.

TL;DR

• Assured Communications is supporting the design, engineering and construction of a new cable landing station (CLS) at Digital Realty’s LAX12 data center in Los Angeles.
• The project brings Assured Communications’ subsea infrastructure expertise to a new cable landing station designed to strengthen connectivity between Asia Pacific and North America.
• The CLS will bring subsea landing infrastructure closer to terrestrial networks, cloud providers, carriers and data center infrastructure.
• Currently in development, the new facility is expected to begin supporting subsea cables in 2028.

# # #

Thousands of miles of ocean separate Asia Pacific and North America, but the digital infrastructure connecting the two regions ultimately comes down to a critical transition point: where subsea cables reach land and connect to the terrestrial networks that carry data onward.

In Los Angeles, Assured Communications is helping build a new piece of that global connectivity ecosystem.

The subsea infrastructure specialist is supporting the design, engineering and construction of a new cable landing station at Digital Realty’s LAX12 data center. The project is designed to establish a new landing point for subsea cable systems while bringing subsea infrastructure closer to the terrestrial networks serving the Los Angeles market.

For Assured Communications, the project puts its experience in subsea cable systems and landing infrastructure development to work on infrastructure designed to support connectivity across the Pacific.

“Assured Communications is supporting the design, engineering and construction of the project, drawing on its experience in subsea cable systems and landing infrastructure development. Projects like this are reshaping how global infrastructure is deployed and connected. We’re excited to help bring a new subsea cable landing station to Los Angeles to further strengthen connectivity between Asia Pacific and North America,” said Joel Ogren, Founder and Chief Executive Officer of Assured Communications.

For subsea cable operators, the new CLS is designed to provide a turnkey solution for landing and powering their systems while providing access to Digital Realty’s broader Los Angeles ecosystem. That integration will provide access to terrestrial networks, cloud providers and carriers across the company’s Los Angeles campus.

By bringing these resources together, the CLS will help create closer integration between subsea and terrestrial networks, giving carriers, cloud providers and content platforms additional routes and greater flexibility to connect between Asia Pacific and North America.

“This development reflects our continued focus on expanding Digital Realty’s role in supporting global connectivity,” said Andy Power, President and Chief Executive Officer of Digital Realty. “By bringing together subsea landing infrastructure with our existing interconnection and data center platform, we are strengthening the intercontinental connectivity available across our Los Angeles campus and providing our customers with a more seamless and scalable environment for connecting across regions.”

With the project currently in development, the new CLS is expected to begin supporting cables in 2028.

The Los Angeles project is another example of Assured Communications putting its subsea infrastructure experience to work, supporting the development of the landing infrastructure needed to connect subsea cable systems with terrestrial networks. As the project advances toward supporting cables in 2028, Assured will continue supporting the design, engineering and construction of the new CLS.

Additional details are available in the full announcement: globenewswire.com/news-release/2026/09/22/3366954/0/en/digital-realty-to-develop-new-cable-landing-station-in-los-angeles.html 

The post Building the Next Subsea Gateway Between Asia Pacific and North America appeared first on Data Center POST.

TL;DR

• AlixPartners’ 2027 Data Center Market Outlook finds that 83% of respondents expect AI inference to require silicon architectures different from today’s GPUs, while 86% of operators anticipate moderate-to-major facility upgrades, redesigns or replacements.
• Near-term market strength is accompanied by uncertainty: 70% of respondents report moderate or strong pricing power, but 64% have reliable demand visibility extending only 12 months or less.
• Rising costs, technology changes and financing challenges are expected to drive industry consolidation, with 62% of respondents viewing data center M&A as attractive now or within the next 12 months.

# # #

The data center industry is entering a pivotal year as artificial intelligence requirements, infrastructure constraints and regulatory pressures reshape the market. According to AlixPartners’ 2027 Global Data Center Market Outlook, demand continues to outpace supply, but limited visibility and rapid technology changes are creating new risks for operators and investors.

The report draws on responses from more than 480 senior executives across the data center ecosystem in North America, Europe, and the Middle East. Participants represented builders, operators, AI infrastructure companies, investors, and lenders working across hyperscale, colocation, edge, enterprise, and neocloud environments.

One of the report’s key findings is the expected transition from AI training to inference workloads. Eighty-three percent of respondents believe inference will require silicon architectures different from today’s GPUs, while 86% of operators anticipate moderate-to-major upgrades, redesigns or replacements of existing facilities. Higher rack densities, direct liquid cooling, and new network configurations could make this transition especially challenging for operators with older infrastructure or limited capital.

Although 70% of respondents currently report moderate or strong pricing power, confidence in future demand remains limited. Sixty-four percent have reliable demand visibility extending only 12 months or less, while 60% believe a major market disruption is extremely or very likely within the next two years. This combination leaves the industry balancing near-term strength against longer-term uncertainty.

Financial pressure is also expected to increase. Sixty-four percent of respondents anticipate more distressed data center situations during the next 12 to 18 months, driven primarily by rising energy costs, insufficient operating cash flow, technology disruption, and competition. Smaller operators may face the greatest risk due to reduced pricing power, debt-maturity challenges, and limited ability to fund necessary upgrades.

Construction and operational challenges will remain central in 2027. Project coordination and construction delays were each cited by 66% of respondents as leading capital deployment challenges, while regulatory restrictions and local community opposition ranked as the top financing obstacle. The findings suggest that operators capable of modernizing their facilities, improving operational efficiency, and maintaining flexibility will be best positioned to navigate the market ahead. Explore AlixPartners’ 2027 Global Data Center Market Outlook for deeper insights into the trends, risks and opportunities shaping the data center market in the year ahead. 

The post AI Demand, Infrastructure Constraints Set the Stage for a Transformative 2027 Data Center Market appeared first on Data Center POST.

TL;DR

• Queue Totals Reflect Intent, Not Delivery: Alberta’s interconnection queue has reached 19,565 megawatts, measuring speculative demand rather than actual capacity being built
• Grid Allocations Are Strictly Constrained: Initial connection allocations are fully committed, leaving developers without existing approvals unable to obtain credible energization dates while permanent frameworks are finalized
• Priority Favors “Bring-Your-Own-Generation”: Under Alberta’s Data Centre Regulation, queue priority is granted to projects that pair data loads with new dedicated generation or storage
• Behind-the-Meter Power Bypasses Queue Delays: Co-located, off-grid self-supply under AUC Rule 007 avoids interconnection requests entirely, trading unmanageable queue delays for manageable execution risks

# # #

By Jolie Kahn, Chief Executive Officer and Director, FingerMotion, Inc.

Alberta has become the most quoted jurisdiction in North American data center development, and the figure doing most of that quoting is the interconnection queue. As of July 30, 2026, the Alberta Electric System Operator reported approximately 19,565 megawatts of requested data centre load awaiting connection. The number is generally presented as evidence of a boom.

It is worth setting that figure beside another one. Alberta’s all-time peak demand — every home, business and industrial facility in the province, drawing at once — is roughly 12,800 megawatts. The queue is more than one and a half times everything Alberta has ever consumed at a single moment.

A number that large is not a forecast of supply. It is a measure of how many parties want the same scarce thing.

A Queue Measures Demand, Not Delivery

Entering an interconnection queue is inexpensive relative to building anything. A queue position commits no one to construction, no one to an offtake agreement and no one to a capital plan. It is closer to an option than a project.

This matters because queue totals are increasingly cited as though they describe capacity that is on its way. They do not. They describe intent, and intent in a constrained market is abundant. Operators evaluating where to site workloads over the next three years are better served by asking a narrower question: how much capacity in this jurisdiction is permitted, and how much is actually energized.

The First Allocation Is Already Spent

Alberta’s answer to the volume of requests was to limit them. Phase 1 of the AESO’s approach permitted 1,200 megawatts of new data centre connections, and that allocation is now fully committed. Phase 2, the permanent framework, is still being finalized. An interim bridging approach allows limited additional load under strict conditions, capped system-wide at 1,600 megawatts, limited to three years per project, and requiring that data centres reduce demand first when the grid is stressed.

The practical consequence is straightforward. A developer without an existing allocation cannot presently provide a credible grid energization date. Not a pessimistic one — none at all, because the framework that would produce it is not yet final.

The Queue Sorts By Who Brings Power

Alberta’s Data Centre Regulation, effective June 2026, formalized the tiebreaker: projects that bring dedicated new generation or storage are prioritized in the connection queue. The province’s bring-your-own-generation framework asks a developer to pair new generation with the load it intends to draw. Meta’s Sturgeon County campus is the widely cited example, combining a grid connection with a multi-billion-dollar private generating station.

This reframes the planning problem. The binding constraint on AI infrastructure in a jurisdiction like Alberta is not land, and it is not hardware. It is permitted power. Capital and equipment can be procured on commercial timelines. A connection cannot be procured at all when the allocation is closed.

Below The Threshold, A Different Path Exists

There is a second route, and it is less discussed because it does not appear in queue statistics at all.

Alberta Utilities Commission Rule 007 governs facility applications for power plants, substations, transmission lines and industrial system designations. Smaller individual power plants fall into a lighter review than large facilities. An Industrial System Designation permits co-located generation and load to be treated as a single industrial system for self-supply purposes, so that internal wiring is not regulated as transmission.

A facility developed this way generates and consumes power on site. It makes no interconnection request, holds no queue position and requires no transmission service arrangement. Because each application is comparatively small, several sites may be permitted in parallel over a period measured in months rather than in the years an interconnection can require.

That is a meaningful difference in a market where the grid path currently offers no date at all.

The Trade Is Real

Behind-the-meter generation is not a shortcut, and it should not be described as one.

Generation capital that the utility would otherwise carry moves onto the developer’s balance sheet. Fuel supply must be contracted rather than assumed. Reliability becomes an operating responsibility rather than a purchased service, and a site without a grid connection has no system to lean on during an outage. Permitting itself is not guaranteed; municipal processes, landowner agreements and regulatory review all consume time, and timelines slip.

The honest characterization is that this path trades one set of risks for another. It substitutes execution risk, which a capable operator can manage, for queue risk, which no operator can manage at all.

What To Ask Instead

The industry would be better served by reporting Alberta’s numbers in four categories rather than one: requested, allocated, permitted and in service. Those are very different quantities, and at present only the first is widely quoted.

For anyone sitting capacity in a constrained jurisdiction, the more useful question is not how large the queue has grown. It is whether a given project needs the queue at all.

# # #

About the Author

Jolie Kahn is Chief Executive Officer and Director of FingerMotion, Inc. (Nasdaq: FNGR). She has been a public-company CEO, CFO and corporate finance attorney for more than three decades, with over $5 billion in capital raises and experience spanning data centers, power, capital markets, fintech and high-growth technology. She previously served as Principal Outside Counsel and General Counsel to a major Bitcoin mining company, overseeing SEC compliance, governance, financings and regulatory matters. She holds a BA from Cornell University and a J.D. magna cum laude from Benjamin N. Cardozo School of Law.

Jolie Kahn has also written on modular data center economics for the Forbes Business Council.

The post A Queue is Not a Pipeline: What Alberta’s Data Center Numbers Actually Say appeared first on Data Center POST.

TL;DR

• AI introduces an operational scaling challenge rather than a raw storage capacity issue, as continuous data flows across training, inference, backup, and archival overwhelm traditional, isolated storage systems.
• Managing fragmented infrastructure has become the primary bottleneck, with separate tools, security policies, and unpredictable AI workloads creating massive operational overhead.
• Autonomous data infrastructure unifies these workloads into a single namespace, automating data placement, lifecycle optimization, and cyber resilience directly at the storage layer.
• IT professionals can transition from routine maintenance to strategic architecture, focusing on governance, security, and system design while allowing adaptive platforms to handle operational tasks autonomously

# # #

By Erwan Girard, Chief Product Officer at Scality

Enterprise infrastructure has always evolved in response to scale. Virtualization addressed server sprawl. Cloud computing reduced the need to provision physical resources for every new application. Automation helped administrators manage environments that had grown too large to operate manually.

Artificial intelligence presents a different kind of scaling problem.

The discussion around enterprise AI has largely centered on models, GPUs, and inference performance, but those technologies represent only a fraction of what organizations must operate. Every production AI deployment creates a continuous flow of data that must be ingested, protected, moved, analyzed, retained, governed, and eventually archived. Those activities place demands on infrastructure that are very different from the workloads storage systems were originally designed to support. 

This is becoming increasingly apparent as organizations move beyond pilot projects. AI is no longer a single workload running on isolated infrastructure. A single application may include high-speed storage for model training, object storage for inference data, lower-cost capacity for operational datasets, immutable storage for cyber resilience, and long-term archives to satisfy regulatory requirements. Traditionally, those functions have been handled by separate products with separate management tools, security policies, and operational teams. That architecture worked reasonably well when data moved slowly and applications followed predictable lifecycles. AI changes both assumptions. 

Training datasets expand continuously. New models are introduced far more frequently than traditional enterprise applications. Inference workloads fluctuate as demand changes. The same dataset may move repeatedly between active processing, backup, compliance, and archival over its lifetime. Each transition introduces another operational task, another opportunity for inconsistency, and another point where administrators must intervene. Before long, the effort required to manage the infrastructure begins to rival the effort required to build the AI applications themselves.

Complexity Has Become the Real Infrastructure Challenge 

For years, the answer to operational complexity was automation. Administrators automated provisioning, scripted maintenance, and orchestrated repetitive tasks. Those capabilities remain valuable, but they were designed to execute predefined actions under predefined conditions. AI environments are considerably less predictable. Infrastructure must continually adapt to changing workloads, shifting performance requirements, evolving security policies, and rapidly growing data volumes, often without the benefit of stable operating patterns.

That is where autonomous data infrastructure represents something more substantial than another automation framework. Rather than treating storage as a collection of independent systems, it starts with the assumption that the platform itself should continuously optimize how data is managed throughout its lifecycle. Capacity, performance, protection, and cost become policy decisions rather than infrastructure projects. Data moves between performance tiers automatically according to business requirements instead of being exported, migrated, and re-imported into separate platforms. A single namespace spans workloads that historically required multiple storage systems, allowing infrastructure to evolve without repeatedly forcing administrators to redesign the environment. 

That architectural change may ultimately prove more important than the automation itself. Many organizations underestimate how much operational complexity accumulates simply from running multiple storage platforms. Every environment has its own authentication model, monitoring tools, lifecycle policies, upgrade schedules, recovery procedures, and performance characteristics. As AI expands across the enterprise, those management layers multiply alongside the data. Reducing the number of operational boundaries often creates greater long-term value than introducing another layer of orchestration.

The same principle applies to cyber resilience. AI has increased the value of enterprise data far beyond traditional business records. Training datasets, model checkpoints, vector indexes, and inference pipelines have become strategic assets in their own right. Protecting them requires more than backup software. It requires infrastructure that assumes failures and attacks will occur and is designed to recover without depending on manual intervention. 

The conversation also extends beyond security to control. As AI initiatives become more strategic, organizations are under growing pressure to understand where data resides, who can access it, and which legal and regulatory frameworks govern it. That is especially true for enterprises operating across multiple countries or in highly regulated industries, where data residency requirements, digital sovereignty initiatives, and industry-specific compliance obligations increasingly influence infrastructure decisions. Rather than treating these as separate governance exercises, modern infrastructure must make location, retention, and access policies part of the data lifecycle itself, enabling organizations to meet regulatory requirements without introducing additional operational complexity.

One of the more significant design decisions behind autonomous data infrastructure is that immutability exists within the storage engine itself rather than being implemented solely through administrative policy. Instead of modifying existing data in place, new versions are written separately while previous versions remain intact. Combined with distributed self-healing that rebuilds only affected objects instead of entire disks, this creates a fundamentally different operational model for resilience. Recovery becomes part of normal system behavior instead of an exceptional event requiring administrators to coordinate lengthy repair efforts. 

From Infrastructure Operators to Infrastructure Architects

Perhaps the most interesting implication has little to do with storage technology itself. Infrastructure teams are already responsible for environments that are growing faster than headcount, and AI is accelerating that imbalance. The objective is not to remove people from operations, but to reduce the amount of time highly skilled engineers spend on repetitive maintenance that adds little strategic value. As more routine activities become policy-driven and continuously optimized, infrastructure professionals can devote more attention to architecture, governance, capacity planning, and aligning technology decisions with business priorities.

That evolution mirrors what is happening across software engineering, networking, and cybersecurity. AI is steadily shifting human expertise away from repetitive execution and toward system design, governance, and strategic decision-making. Autonomous data infrastructure reflects the same progression. Rather than asking administrators to manage an ever-growing collection of storage products, it treats the infrastructure as an adaptive system that operates within policies established by the people responsible for it. The most effective approach is not to take humans out of the loop, but to keep them in control of the decisions that shape security, compliance, and business outcomes while allowing the platform to execute routine operational tasks autonomously. 

Viewed from that perspective, autonomous data infrastructure is less about storage than it is about preparing enterprise IT for the next decade. AI has exposed the limitations of architectures built around isolated products, manual coordination, and steadily increasing operational overhead. Organizations will continue investing in faster GPUs and more capable models, but those investments will deliver their greatest value only if the infrastructure beneath them becomes equally capable of managing complexity. The next generation of enterprise infrastructure will not simply store data more efficiently. It will actively participate in operating the environments that modern AI depends upon.

# # #

About the Author

Currently serving as the Chief Product Officer at Scality, Erwan Girard has been instrumental in shaping the company’s product strategy and driving its success. In his tenure, he has built an all-round product management team focused on serving and bringing value to Scality customers and prospects. He has led RING product evolutions and launched Artesca in order to better address the Object storage market resulting in significant revenue growth and market expansion.

The post AI Didn’t Create a Storage Problem. It Created an Operations Problem. appeared first on Data Center POST.

TL;DR

• Redefining the Assessment Boundary: While data centres carry a significant operational and lifecycle footprint, judging them solely at the facility boundary creates an incomplete environmental evaluation that ignores what the digital infrastructure enables, reduces, or replaces.
• Measuring Net System Impact: A complete evaluation requires a three-part ledger accounting for the direct lifecycle burden, enabled environmental benefits (avoided physical journeys, energy savings, extended asset life), and potential rebound demand.
• Testing Structural Displacement: The Structural Displacement Ratio measures whether enabled benefits, after subtracting rebound effects, exceed the direct lifecycle burden, requiring verifiable evidence rather than assumed environmental credit.
• Demanding Independent Assurance: Digital infrastructure is not inherently sustainable; policymakers, investors, and operators must mandate independent assurance and clear evidence quality controls across the entire system boundary

# # #

A complete environmental ledger must measure direct burden, enabled benefit and rebound

The environmental cost of data centres is real, but a facility boundary is not a system boundary. Decision makers need evidence of the physical activity digital infrastructure enables, reduces or replaces, alongside an uncompromising account of its own lifecycle burden.

The data centre has become one of the most visible symbols of the environmental cost of the digital economy. Electricity demand is large and concentrated. Cooling can place pressure on water systems. Construction consumes land, concrete, steel and electrical equipment, while servers rely on complex supply chains and short replacement cycles.

Those impacts must be measured without euphemism. Yet visibility is not the same as causation, and measuring only what enters the facility can produce an incomplete environmental judgement. A data centre does not exist for its own sake. It processes information used by buildings, utilities, transport systems, businesses, public services and consumers. The relevant question is therefore not simply what the facility consumes. It is what the complete digital system consumes, enables, reduces and replaces.

The wrong boundary

A kilowatt hour consumed in a server hall is immediately visible. The effect of a digital service elsewhere in the physical economy is usually dispersed. Energy saved through intelligent building controls may be distributed across thousands of assets. Water preserved through network leakage detection appears through many small interventions. Avoided journeys, more efficient logistics and longer equipment life are rarely attributed back to the digital infrastructure that enabled them.

This creates an asymmetric ledger. The direct burden is counted at the facility boundary, while possible reductions in physical consumption remain outside it. The answer is not to award data centres an assumed environmental credit. It is to widen the system boundary and apply the same evidential discipline to both sides.

Three ledgers and an evidence test

A practical Net System Impact assessment begins with three distinct ledgers. The first records the direct lifecycle burden: site, construction, embodied carbon, operational electricity, cooling, water, backup generation, network infrastructure, hardware manufacture, replacement, electronic waste, decommissioning and reuse.

The second record enabled environmental benefits. This may include physical activity demonstrably reduced through digital services, such as avoided travel, optimised energy use, prevented water loss, reduced material consumption or extended asset life. These benefits must be tested against a credible counterfactual. What would have happened without the digital intervention, and would an equivalent outcome still have been required?

The third records rebound and additional demand. Efficiency can lower cost and increase convenience, causing total consumption to rise. New digital capability can also create activity that did not previously exist. These effects cannot be hidden inside a net figure.

A fourth control sits across all three ledgers: evidence quality. The boundary, baseline, causal relationship, allocation method and assurance status must be visible. Where evidence is weak, the claimed benefit should be discounted rather than promoted as fact.

The structural displacement test

The proposed Structural Displacement Ratio asks whether the enabled benefit, after rebound, exceeds the direct burden. In simple terms, the test is: Structural Displacement Ratio = enabled benefit minus rebound, divided by direct lifecycle burden.

A result above one may indicate that the system eliminates more physical waste than it creates, but only within the environmental category being tested and only when the evidence is credible. A carbon result cannot erase a material failure involving water, land, biodiversity, resilience or community impact. Nor should a favourable calculation become a permanent entitlement. 

Counterfactuals change as grids decarbonise, regulations tighten and alternative technologies improve. The assessment must therefore expire and be retested.

Technology supports proof but does not create it

Internet of Things devices can connect environmental claims to measured physical conditions. Artificial intelligence can interpret complex data and optimize performance. Blockchain can preserve provenance, permissions, attribution and retirement across organisations. None of these technologies creates environmental value independently.

Value arises when trusted information changes physical outcomes. A smart building that gathers more data but consumes the same resources has not demonstrated an environmental benefit. A predictive maintenance system matters when evidence shows that it prevented failure, avoided replacement or extended useful life. A digital logistics platform matters when it reduces real journeys, fuel or wasted capacity against a defensible baseline.

A decision framework for owners and regulators

Data-centre planning and investment decisions should therefore require a complete lifecycle account and a defined statement of the physical systems the proposed workloads are expected to influence. 

Operators and customers should identify measurable outcomes, allocate benefits conservatively and prevent double counting between the data centre, software provider, asset owner and end user.

Independent assurance is also essential. The party benefiting from a sustainability claim should not be the only party defining the counterfactual or verifying the result. Evidence should be capable of review by investors, regulators, customers and affected communities without disclosing commercially sensitive operational data unnecessarily.

The conclusion is deliberately conditional. Digital infrastructure is not inherently sustainable, and scale alone will not make it so. Poorly located or weakly governed facilities can deepen pressure on grids, water systems and communities. Equally, treating the data centre as an isolated consumer can ignore the inefficient physical activity that digital services may displace.

The data centre is not the real white whale. The real target is the accounting boundary that separates what digital infrastructure consumes from what it replaces. Society should neither pursue data centres as convenient symbols nor accept the industry’s benefits without proof. It should demand a complete ledger.

Full paper

The complete TUTUM white paper, Is the Data Centre the Next White Whale?, including the full Net System Impact Assessment and illustrative calculations, is available here. 

# # #

About the Author

Mark Burns is Executive Director of TUTUM Group and a chartered development and infrastructure leader with nearly three decades of experience across sustainable master development, real estate, infrastructure and complex programme delivery. He previously served as Senior Development Director at Masdar City and has delivered six LEED Platinum projects client-side. His work focuses on asset governance, programme assurance, sustainability and the relationship between physical infrastructure and digital capital.

The post The Data Centre Should Be Judged by What It Replaces appeared first on Data Center POST.

TL;DR

• Major local investment: Empire Fiber Internet has launched 100% fiber internet in Berwick, PA, supported by a $1.4 million investment and now available to more than 2,800 homes and businesses.
• New choice for connectivity: The build gives Berwick residents and local employers access to symmetrical fiber speeds, responsive local support and service plans up to 2 Gig.
• Momentum across the region: The Berwick launch builds on Empire Fiber Internet’s continued growth across Pennsylvania and New York, including recent expansions in the Lehigh Valley, Steuben County, Wellsville, Ithaca, and Victor.

# # #

Empire Fiber Internet, a leading fiber optic internet service provider serving communities across New York and Pennsylvania, has launched its 100% fiber network in Berwick, PA, bringing fast, reliable internet service to more than 2,800 homes and businesses.

Enabled by a $1.4 million local investment, the new network gives the Columbia County community another choice for the high-capacity connectivity increasingly essential to daily life and local commerce.

For Berwick, the investment comes at an important time. High-speed broadband has become foundational infrastructure for families, schools, employers and small businesses alike as it enables remote and hybrid work, online learning, telehealth, cloud-based business applications, e-commerce and customer communications. As Berwick continues to support local employers, attract investment and meet the digital needs of residents, the availability of a new fiber network strengthens a critical piece of the community’s infrastructure.

“Reliable, high-speed internet is essential infrastructure for today’s businesses, whether it’s a manufacturer moving large amounts of data, a small business processing transactions, or an employer simply trying to stay connected with customers and employees,” said Chris Berleth, President/CEO of the Columbia Montour Chamber of Commerce. “Empire Fiber Internet’s investment in Berwick brings another broadband option to the community, and that kind of private investment and increased choice helps strengthen the environment in which our local businesses operate and grow.”

This latest launch also reflects Empire Fiber Internet’s broader expansion strategy across Pennsylvania and New York. In 2026 alone, the company announced new or expanded fiber availability in Pennsylvania’s Lehigh Valley and in New York communities including Steuben County, Wellsville, Ithaca, and Victor.

To learn more about Empire Fiber Internet, visit empirefiber.com.

The post Empire Fiber Internet Connects More Than 2,800 Berwick Homes and Businesses appeared first on Data Center POST.

TL;DR

• • Community concerns about data centers often extend beyond infrastructure to questions about AI, resource use, surveillance and Big Tech.
  • Effective community engagement starts with understanding local priorities rather than immediately presenting a project’s benefits.

• Communities need clear information about potential impacts, including power, water, noise, land use and economic development.
• Engagement strategies should reflect the priorities and characteristics of each individual community.
• Sharing community engagement lessons across the industry could help improve how future projects are introduced and developed.

# # #

Data center development is accelerating to support growing demand for cloud computing, artificial intelligence and other digital services. At the same time, projects are facing increased public scrutiny as communities ask more questions about power, water, land use, environmental impacts and what large-scale digital infrastructure development means for them.

A recent Cool Vector podcast episode, “Data Centers Have a Community Engagement Problem,” explored this changing environment with Megan Baker, Founder and Principal of Tuli Public Affairs; Ilissa Miller, Founder and CEO of iMiller Public Relations (iMPR); and Phillip Koblence, Co-Founder of Nomad Futurist and Head of Critical Ventures. Moderated by Cool Vector host David Snow, the conversation examined why community engagement needs to become a more integral part of data center development.

Understanding What Is Driving Community Concerns

For decades, much of the public had little reason to think about the physical infrastructure supporting digital services. Today, data centers are increasingly associated with AI, energy demand, water consumption, Big Tech and other issues receiving significant public attention.

“People fundamentally fear what they are not comfortable with,” said Koblence. That uncertainty can make technical explanations alone insufficient when residents have broader concerns about how development could affect their communities.

Miller emphasized the emotional component behind some of those concerns. “It’s all rooted in fear and whether that fear is justified or not is secondary, but it’s a real emotion,” she said. “And so that’s why the facts don’t matter, because it’s emotional. They’re afraid of AI. They’re afraid of surveillance.”

Addressing those concerns requires understanding the questions behind them. Communities need information about potential impacts, but they also want to understand who is developing the project, why it is coming to their community and how local priorities will be considered.

Start With Listening

Rather than beginning with tax revenue, investment, or other anticipated project benefits, the discussion emphasized listening to communities before developing an engagement strategy. Understanding local priorities can help shape how developers communicate and identify issues that may need to be addressed earlier in the process.

“They have to slow down, listen,” said Miller. She later summarized the approach simply: “Go to these communities with curiosity.” This community-first approach puts greater emphasis on understanding what residents and local stakeholders want, need, and value before determining how a project should be communicated.

No two communities are the same, which makes a one-size-fits-all approach to community engagement ineffective. Some communities may prioritize economic development, while others may be more focused on natural resources, preserving community character or managing growth.

“We need to recognize too that each community is so different, right?” said Baker. “It can’t be a blanket procedure of going into a community and just deploying the same tools or the same process over and over again. You need to spend the time to understand, what does this community value?”

Community research and stakeholder engagement conducted early in the development process can help answer those questions before a project becomes the subject of public debate. It can also provide local leaders with better information when residents begin asking questions about power, water, noise, jobs, taxes and other potential impacts.

Sharing Lessons Across the Industry

Community engagement in one market can also influence perceptions elsewhere. Residents may view controversial projects as representative of the broader data center industry rather than the actions of a single developer, making community relationships an industry-wide concern.

Baker argued that companies have an opportunity to learn from one another rather than treating engagement strategies as proprietary. “The community engagement and sustainability tactics of these projects aren’t proprietary and they shouldn’t be. And we should be more collaborative about sharing the things that are going well and the things that aren’t going well so that we can all be moving forward with less friction,” she said.

The podcast also highlighted the Digital Infrastructure Framework being developed through OIX, which is intended to provide communities with guidance for understanding and evaluating digital infrastructure development. Efforts such as these could give municipalities additional tools for asking questions while helping developers better understand what communities need from them.

As data center development continues, community engagement is becoming another important consideration alongside power, land, connectivity and permitting. Starting earlier, listening before communicating project benefits and approaching each community individually can help developers build stronger relationships with the people and places that will ultimately host digital infrastructure.

To hear the full conversation with Megan Baker, Ilissa Miller and Phillip Koblence, listen to Cool Vector’s “Data Centers Have a Community Engagement Problem” on Substack here.

The post Data Centers Have a Community Engagement Problem: Why Listening Must Come First appeared first on Data Center POST.

TL;DR

• Physical constraints dictate deployment viability: Rack power density, liquid cooling capabilities, and grid interconnection timelines now determine AI deployment success rather than traditional floor space or total facility square footage.
• Avoidable sequencing and architecture mistakes: Enterprise buyers frequently purchase high-powered accelerators before verifying per-rack power or cooling constraints, or misallocate budgets by buying training-class clusters for inference workloads.
• Low utilization is the dominant cost leak: Unscheduled idle capacity, overprovisioned inference clusters, and unmonitored data egress fees multiply effective costs far more than hardware unit prices.
• Strategic preparation for 2026: Organizations must audit legacy colocation contracts, price data exit costs, and instrument actual utilization metrics before committing to new hardware procurement

# # #

The AI data center infrastructure conversation has been dominated by the supply side: megawatts secured, land acquired, cooling retrofitted, grid queues navigated. Less attention has gone to the demand side, where enterprise buyers are making commitments they do not fully understand.

That gap is becoming expensive. Organizations are signing capacity agreements, buying accelerators, and choosing platforms based on assumptions that stopped being true two hardware generations ago. Operators and providers see the results: customers arriving with hardware their contracted space cannot power, or with clusters that run at a fraction of their capability.

This piece looks at what has actually changed at the facility level, and where buyers on the other side of the table are getting it wrong.

What Has Actually Changed in AI Data Center Infrastructure?

The changes that matter are physical rather than digital. Rack power density, heat removal, and grid interconnection timelines now determine what is possible, and none of them can be resolved through procurement or software.

Three shifts sit underneath everything else.

Power density outgrew existing facility design. Halls provisioned for general enterprise computers were built around assumptions about per-rack draw that AI accelerator deployments break. Total facility capacity has become a less useful figure than what a single rack position can deliver.

Air cooling reached a practical ceiling. Beyond a certain density, the airflow and volume required make air cooling uneconomical before it becomes technically inadequate. Liquid has moved from specialist option to default assumption for dense AI deployments.

Power availability became the site selection criterion. In multiple markets, interconnection timelines now exceed construction timelines. That has pushed site selection toward available electricity and driven interest in on-site generation and long-term power purchase agreements.

For the industry, none of this is new. For the enterprise buyer signing a three-year agreement, most of it still is.

Why Rack Density Has Become the Deciding Specification

Rack-level power availability is now the specification that determines whether a deployment is viable. Facility totals and floor area say little about whether specific AI hardware can be installed and operated in a given position.

This is where the most common buyer failure occurs. A customer evaluates space on square footage and total capacity, signs, and then discovers that the contracted kilowatts per rack falls short of what their accelerator configuration draws. The hardware is purchased. The space is contracted. Neither can be used as planned.

Providers increasingly see this arrival as an escalation rather than a question. The underlying problem is that legacy procurement habits carried into a workload category they were never designed for.

The practical guidance for buyers is narrow and specific: treat power per rack as a contractual figure, confirm it in writing, and establish whether the position supports the density today or only after a future phase of work.

How Cooling Moved from Utility to Design Constraint

Cooling has shifted from a background consideration to a primary constraint on deployment. Most new high-density AI capacity assumes some form of liquid cooling, and retrofit programs are widespread across existing estate.

The point buyers most often miss is that liquid cooling is a facility decision, not an IT decision. It involves floor loading, plumbing routes, leak detection, maintenance access, and, in leased environments, the landlord. Enterprise teams that select accelerators before establishing cooling capability create a sequencing problem that takes months to unwind.

Water use has also moved up the agenda. Cooling designs vary considerably in consumption, and in water-stressed regions this has become a permitting and community-relations matter. Water usage effectiveness (WUE) now appears alongside PUE in enterprise supplier questionnaires with increasing frequency.

Why Networking and Storage Are Part of the Compute Decision

AI training clusters generate heavy east-west traffic between nodes rather than the north-south pattern typical of enterprise applications. The interconnect stops being a supporting system and becomes part of the compute platform itself.

In distributed training, nodes synchronize constantly. A slow fabric leaves expensive accelerators waiting, which is why these environments use InfiniBand or high-speed RDMA over Converged Ethernet rather than standard enterprise networking.

Storage follows the same logic. Training reads large volumes continuously, and a tier that performs adequately for business applications will starve a cluster. Bandwidth matters more than capacity, which is why NVMe tiers and parallel file systems sit in front of colder object storage.

There is a third factor that shapes buyer decisions more than either, and it is rarely priced during evaluation. Data gravity constrains architecture over time. Computers are comparatively portable; large datasets are not. Once training data settles in a particular provider or region, moving it becomes slow and expensive, which quietly narrows every future platform option.

Buyers evaluate computer pricing carefully. Very few price the exit.

Training or Inference: The Distinction Most Buyers Miss

Most enterprise organizations need inference capacity, not training capacity. Training builds or fine-tunes models and demands large, tightly coupled clusters. Inference runs finished models and is latency-sensitive, distributed, and far less demanding per request. Conflating the two drives a substantial share of misdirected spending.

This distinction has commercial consequences for providers as well as buyers. An enterprise customer asking about training-class capacity when their actual requirement is regional inference is a customer who will either overbuy and churn, or discover the mismatch during deployment.

The qualifying question is simple and rarely asked early enough: is this organization building models, or deploying them?

Build, Collaborate, or Rent?

For most enterprise buyers, renting remains the correct default. Ownership makes sense at sustained, predictable utilization. The deciding variable is workload stability, not unit pricing.

Ownership converts a variable cost into a fixed one. That is an advantage when demand is stable and a liability when it is not and enterprise AI demand in 2026 is still, for most organizations, unstable.

Where AI Infrastructure Budgets Actually Leak

The largest avoidable cost in AI infrastructure is low utilization, not unit price. Idle accelerators cost what busy ones cost. A cluster running well below capability multiplies effective cost per useful hour, and no negotiated rate compensates for it.

The recurring leaks are consistent across organizations:

• Idle capacity between workloads, where provisioning matches peak demand but scheduling does not exist
• Overprovisioned inference, sized permanently for spiky traffic
• Data movement, where egress and cross-region transfer are underestimated during evaluation
• Duplicate environments, with development and experimentation clusters running continuously
• Model inefficiency, where a larger model than the task requires becomes an infrastructure cost

The first meaningful saving in most organizations comes from scheduling and right-sizing rather than from buying differently. That is an uncomfortable message for a procurement cycle, but it is generally the accurate one.

A Preparation Checklist for 2026

Preparation is mostly a sequencing problem. The technical decisions are less difficult than the order in which they are made.

1. Classify workloads before selecting a platform. Separate training, fine-tuning, and inference, with volume and latency requirements for each.
2. Audit existing contracts. Check power per rack, cooling capability, and expansion rights in agreements signed before AI hardware became a requirement.
3. Treat power and equipment lead times as project constraints, not procurement details.
4. Price the exit. Model the cost of moving data and workloads to another provider before committing.
5. Instrument utilization before expanding. Accelerator utilization, queue depth, and cost per inference should be measured, not estimated.
6. Plan for shorter refresh cycles. Accelerator generations are moving faster than traditional depreciation models assume.
7. Establish the compliance position early. Where data is processed and moved affects obligations under GDPR, sector frameworks such as HIPAA, and emerging AI-specific regulation.
8. Assess capability alongside capacity. Platform engineering, MLOps, and data engineering gaps account for more failed deployments than hardware decisions do.

Frequently Asked Questions

How much power does an AI server rack need?

Substantially more than a traditional enterprise rack, with the gap widening across accelerator generations. Rather than relying on a general figure, buyers should obtain the specific draw for their configuration from the hardware vendor and confirm in writing that the facility can deliver it per rack position.

Can existing data centers be retrofitted for AI workloads?

Often yes, but the work is structural. Retrofits typically involve electrical distribution upgrades, cooling infrastructure, floor loading assessment, and plumbing. Timelines are driven by equipment lead times and, where applicable, grid connection. A retrofit should be scheduled as a construction project, not an IT refresh.

Is liquid cooling required for all AI workloads?

No. Lower-density AI deployments can still run on air cooling or rear-door heat exchangers. The determining factor is the power density of the specific hardware, which establishes whether air can remove the heat economically. Dense training and high-throughput inference deployments generally require liquid.

Is it cheaper to run AI workloads on-premise or in the cloud?

Neither is universally cheaper. Cloud trends to cost less for variable demand; colocation or ownership tends to cost less at high, sustained utilization. Comparisons that exclude data transfer, staffing, refresh cycles, and exit costs typically overstate the case for ownership.

What is the difference between PUE and WUE?

Power usage effectiveness measures how much of a facility’s total energy reaches IT equipment rather than overhead such as cooling. Water usage effectiveness measures water consumed per unit of IT energy. Both now appear in enterprise supplier assessments, with WUE gaining prominence as cooling water use attracts regulatory attention.

How long does it take to deploy AI infrastructure?

It depends on the route. Cloud capacity can be immediate. Colocation depends on hardware lead times and facility readiness. Building or substantially retrofitting is measured in quarters or years, and is frequently gated by electrical equipment availability and grid connection rather than construction.

How can organizations reduce AI infrastructure costs without reducing capacity?

By addressing utilization before procurement. Scheduling and queueing across teams, right-sizing inference deployments, shutting down idle non-production environments, and selecting appropriately sized models typically recover more budget than renegotiating unit rates.

Key Takeaways

AI has changed data center infrastructure at the physical layer, and the industry has largely absorbed that. The lag is on the demand side, where enterprise buyers continue to apply procurement habits formed around a different class of workload.

The failures are consistent and avoidable. Capacity evaluated on floor area rather than rack density. Hardware selected before cooling capability is established. Training-class infrastructure purchased for what is fundamentally an inference requirement. Utilization treated as an operational detail rather than the dominant cost driver.

For providers, these are recurring conversations worth having earlier in the sales cycle. For enterprise buyers, the most useful preparation for 2026 is not deciding what to buy. It is establishing, with measurement rather than assumption, what the organization actually needs.

# # #

About the Author

Bijal Soni, Co-Founder of Briskstar, is a visionary leader shaping the future of software innovation. With a proven track record of building and leading impactful solutions across web, mobile, and cloud platforms, he blends technical depth with strategic foresight.

The post AI Data Center Infrastructure in 2026: What Enterprise Buyers Need to Prepare For appeared first on Data Center POST.

TL;DR

• Easing AI & Data Center Power Bottlenecks: The technology directly targets growing energy and cooling constraints in AI data centers and mission-critical facilities, allowing operators to expand capacity without adding costly electrical infrastructure.
• Built-In Harmonic Mitigation: By embedding Active Front End harmonic reduction directly into the motor system, Infinitum reduces electrical noise below IEEE 519 standards, lowering thermal load, saving floor space, and boosting reliability.
• Product Momentum & Event Presence: The draft highlights recent product expansions in integrated EC fans and previews Infinitum’s participation at upcoming industry conferences including Yotta, OCP Global Summit, and DCD Connect.

# # #

Data Center POST had the opportunity to connect with Ben Schuler, CEO & Founder of Infinitum. Prior to founding Infinitum, Ben spent the previous 15 years leading the design and deployment of complex industrial automation equipment across Taiwan, Japan and the United States. He was involved in the production ramp of Corning’s Gorilla Glass, Bloom Energy’s solid oxide fuel cell and Solyndra’s thin film solar module.

Infinitum is the direct beneficiary of Ben’s expertise, which helps him lead the mission of disrupting the way industrial machinery and electric motors consume and create electricity. As one of the original creators of Infinitum’s air-core motor system, Ben holds 36 patents related to electric machines and drives.

What does Infinitum do?

Infinitum develops next-generation motor and fan systems that improve power quality and maximize electrical capacity by reducing harmonics at the source, enabling power-constrained facilities to support more critical loads without expanding electrical infrastructure.

What problems does Infinitum solve in the market?

Traditional motors are often heavy, inefficient at variable loads and difficult to service at the end of their useful life. Infinitum is advancing the century-old motor industry with a new design that prioritizes efficiency, reliability, serviceability and responsible material use. This approach delivers meaningful improvements in energy consumption, operational performance and lifecycle impact.

What are Infinitum’s core products or services?

Infinitum’s core products are our Aircore EC+, Aircore EC+ Fans, and Aircore EC+ Pumps. You can learn more about these products on our website, goinfinitum.com/aircoreecplus.

What markets do you serve?

We serve a variety of markets, including data centers and mission-critical facilities, commercial buildings and HVAC, industrial facilities, healthcare, universities and schools, and agriculture.

What challenges does the global digital infrastructure industry face today?

Infinitum’s roots are in sustainability, and its technology is increasingly addressing one of the most urgent challenges facing digital infrastructure: meeting growing power and cooling demands as AI adoption accelerates and available electrical capacity becomes a business constraint.

Organizations are under intense pressure to balance growth, uptime and energy efficiency without adding costly electrical infrastructure or increasing strain on the grid. To help solve this challenge, Infinitum developed the Aircore EC+ motor system, enabling customers to do more with the power they already have.

How is Infinitum adapting to these challenges?

Rather than relying on conventional approaches that use bulky, heat-generating harmonic filters, Infinitum’s engineering team designed unique embedded harmonic mitigation directly into the motor system. This innovation improves power quality while recovering valuable floor space, reducing thermal load and lowering energy losses. By maximizing usable electrical capacity and maintaining high efficiency across the load curve, Aircore EC+ helps facilities increase performance while consuming less energy and requiring less infrastructure.

Equally challenging was proving that sustainability and operational performance can improve together. Infinitum conducted extensive testing, validation and real-world deployments to demonstrate that customers could reduce energy consumption, optimize available power and maintain reliability simultaneously.

The company successfully deployed this technology in some of the world’s most demanding environments, including AI data centers, hospitals and other mission-critical facilities, where performance standards are exceptionally high and downtime is not an option.

What are your company’s key differentiators?

Traditional variable frequency drives (VFDs) are essential for improving motor efficiency, but they produce harmonics that can degrade electrical performance, reduce usable power and threaten uptime. As power demands from AI and electrification rise, harmonics can have significant operational and financial consequences in demanding environments where energy efficiency, uptime and electrical performance are critical.

Aircore EC+ enables customers to do more with less by maximizing available power, electrical performance and equipment life while reducing energy use, footprint and system complexity.

Infinitum’s Aircore EC+ is uniquely designed with integrated Active Front End technology within the motor’s VFD to reduce harmonics at the source and below IEEE 519 specifications, delivering cleaner power without the added cost, footprint, and power losses of conventional passive and active filters. Analysts have commented:

”The Infinitum design delivers a near-sinusoidal waveform with harmonic distortion reduced to well below the IEEE 519 standard, far outperforming traditional passive filtering systems that typically only reduce harmonics to 10%,“ said Giselle Canahuati, Senior Consultant, Frost & Sullivan.

Available and application-specific features include:

• Standard IP65 Enclosure: Heavy-duty exterior protects against dust, dirt and water, allowing motors to operate reliably in harsh outdoor or wet environments.
• Standard High Resistance Ground: Delivers normal operation during an electrical ground fault and reduces the chance of equipment failure. Power-loss ride-through capability enables systems to continue operating during standby power.
• Standard Hybrid Ceramic Bearings: As compared to standard bearings, hybrid ceramic bearings do not conduct electricity and resist corrosion, enhancing durability, reliability and efficiency in high-performance applications.

What can we expect to see/hear from Infinitum in the future?

As customers face power budget constraints related to AI, electrification and higher electricity costs, our product development will continue to focus on helping customers do more with less. That translates into products that require fewer motors, fewer kilowatts and less electrical infrastructure. We’ve continued to innovate and expand our product portfolio for specific applications that help the most energy-intense sectors, including heavy industry, commercial buildings and data centers, to optimize how power is used and conserved. Most recently we recognized customer challenges in the EC fan market, and unveiled a high-efficiency, integrated (fan, motor, variable frequency drive) solution for HVAC, industrial and mission critical applications. We are expanding our ability to deliver more fully integrated systems. With Aircore EC+ Fans, Infinitum optimizes the fan, motor and variable frequency drive together to address real-world power, performance and integration challenges.

What upcoming industry events will you be attending?

Infinitum will be attending a variety of data center industry events, including:

• Yotta
• OCP Global Summit
• DCD Connect Virginia
• 7×24 Exchange Fall Conference

For those interested in connecting during the conferences to learn more about Infinitum’s motor and fan systems for mission critical facilities, please reach out to Anthony Lou at alou@goinfinitum.com or John Garza at jgarza@goinfinitum.com.

Do you have any recent news you would like us to highlight?

• Infinitum Introduces Aircore EC+ with Premium Features to Maximize Available Power Capacity
• Infinitum Unveils New EC Fan Systems
• Infinitum’s Aircore Motor System Earns 2026 E+E Leader Award

Is there anything else you would like our readers to know about Infinitum and capabilities?

AI-driven data center growth, electrification and an aging grid are making power availability a defining constraint. Because motors and fans are embedded throughout cooling infrastructure, improving their efficiency and electrical performance can have a meaningful impact on available power, operating costs and facility capacity. Infinitum helps customers do more with less by delivering more cooling and performance with less energy, material, space and infrastructure.

Where can our readers learn more about Infinitum?

Readers can learn more about us on our website, goinfinitum.com. You can also contact us with any questions using the contact form on our website, found here.

# # #

About Data Center POST

Data Center POST provides a comprehensive view of the digital infrastructure landscape, delivering industry insights into the global data center ecosystem. As the industry’s only peer-contributed and online publication, we offer relevant information from developers, managers, providers, investors, and trendsetters worldwide.

Data Center POST works hard to get the most current information and thought-provoking ideas most apt to add relevance to the success of the data center industry. Stay informed, visit www.datacenterpost.com.

If you are interested in contributing to Data Center POST, contact us at contributions@datacenterpost.com or submit your article here.

Want more digital infrastructure news? Stay in the know and subscribe to Data Center POST today!

The post Company Profile: How Infinitum is Overcoming Digital Infrastructure Power Constraints appeared first on Data Center POST.

TL; DR

• The standard model of grid-dependent power, diesel backup, and brief UPS support is increasingly inadequate for today’s volatile, AI-driven workloads.
• DCD Energy & Power Week 2026 will host a panel on maximizing battery asset utilization across data center facilities.
• The Oct. 1 session covers storage placement, layered storage strategies from grid-scale to rack-level, and chemistry and lifecycle tradeoffs.
• Panelists will also explore the strategic role of BESS in decarbonization through renewables capture and grid intermittency mitigation.

# # #

As AI-driven workloads push data centers toward higher densities and more volatile power demands, the standard model of grid-dependent power, diesel-fueled backup, and brief UPS support is increasingly inadequate for managing today’s environmental mandates and operational pressures. On Oct. 1, on its Energy & Sustainability Channel, DCD will host a broadcast titled “The data center battery revolution” from 9-10 a.m. EDT on its Energy & Sustainability Channel. The conversation will explore how Battery Energy Storage Systems (BESS) are transitioning from simple emergency reserves into core strategic assets that support both operational resilience and grid flexibility.

The streaming session brings together leaders from pioneering battery technology providers to deconstruct how operators can maximize battery asset utilization within data center facilities today. Topics include redefining the role and physical placement of energy storage within the modern mission-critical power train; the commercial use cases and strategic trade-offs for layered storage, from grid-scale assets down to rack-level storage; the density, lifecycle, supply chain, and chemistries of current storage options in the market; and the strategic utility of BESS in advancing decarbonization through the capture of surplus renewables and mitigation of grid intermittency.

Who:

• John Gagge, Vice President of Business Development, Network and Infrastructure Solutions, EnerSys
• Liz Wallace, Senior Product Manager, ZincFive
• Eric Hill, General Manager for Product and Marketing, CSB Energy Technology

What: “The data center battery revolution” panel discussion

When: Thursday, Oct. 1, 9-10 a.m. EDT

For operators, the core issue is practical: As workloads grow more volatile and environmental mandates become more stringent, how should BESS be positioned – physically and strategically – within the power train to deliver both resilience and flexibility? That’s where battery architecture becomes more than a backup decision and starts to shape overall infrastructure and decarbonization strategy.

Whether you’re designing new data center infrastructure, upgrading existing facilities, or evaluating future power strategies, this session offers actionable insight into one of the industry’s fastest-evolving challenges.

To register to stream this discussion, please visit: www.datacenterdynamics.com/en/dcd-broadcasts/energy–power-week-2026/the-data-center-battery-revolution-overcoming-myths-to-secure-future-growth

The post The Data Center Battery Revolution appeared first on Data Center POST.

TL;DR

• Extended Project Scope: A data center technology refresh does not end when new hardware goes live, as retired IT assets and facility infrastructure still require clear disposition plans. Leaving retired equipment in storage leads to lost resale value, wasted space, and tracking challenges.
• Inventory & Security Controls: Retired hardware must be cataloged by serial number and configuration, with data-bearing devices kept under a strict chain of custody for sanitization aligned with standards like NIST SP 800-88.
• Flexible Value Recovery: Organizations can recover significant operational and financial value by assessing assets for internal redeployment, secondary market resale, OEM trade-ins, or parts harvesting. Internal redeployment alone can save up to 85% compared to purchasing new equipment.

# # #

A data center technology refresh does not end when new servers, storage systems, or networking equipment goes live.

The retired equipment still needs a clear next step. Some assets may still have useful life and resale value. Others may contain sensitive data, reusable components, or materials that need to be routed for reuse, resale, recycling, or disposal

If these assets sit in storage, companies can lose resale value, miss internal reuse opportunities, take up valuable space, and make asset tracking more difficult.

In this blog, you will learn what happens to data center assets after a technology refresh, how retired equipment is evaluated, where it can go next, and why planning disposition early can help recover more value.

What Assets Are Retired During a Data Center Technology Refresh?

The assets retired during a data center technology refresh depend on the scope of the project. A server upgrade may remove only a limited set of IT equipment, while a larger refresh can affect storage, networking, power, cooling, and other supporting infrastructure.

The retired assets usually fall into two broad groups.

     1. Data-Bearing and IT Equipment

This group includes equipment that stores, processes, or transmits data, such as:

• Servers
• Storage arrays
• Hard drives and SSDs
• Network switches
• Routers
• Firewalls
• Backup appliances
• Other computing hardware
  

      2. Supporting Data Center Infrastructure

A refresh may also leave behind equipment that does not primarily store data but still has reuse, resale, or material value, including:

• Server racks and cabinets
• Power distribution units
• UPS equipment
• Batteries
• Cabling
• Cooling equipment
• Power and electrical components

What Happens to Data Center Assets After They Are Retired?

Before retired data center equipment is reused, sold, or recycled, it first needs to be identified, secured, and assessed. This creates a clear record of what was removed and helps determine the next step for each asset.

     1. Assets Are Identified and Tracked

Each asset should be recorded using details such as its asset tag, serial number, manufacturer, model, configuration, age, condition, location, data-bearing status, and ownership records.

A complete retirement inventory helps teams confirm what was removed, where it is located, and who is responsible for it. It also supports value recovery by showing which assets may still be suitable for reuse, resale, parts recovery, or another disposition route.

     2. Data-Bearing Assets Are Separated

Servers, hard drives, SSDs, storage systems, and other devices that may contain company data require controlled handling after removal.

These assets should remain under a documented chain of custody and be placed in secure storage until the required media sanitization or destruction process is completed. Teams should also keep records of data destruction and verify that sanitization was completed before the equipment moves to its next destination.

The National Institute of Standards and Technology (NIST) SP 800-88 Rev. 2 provides guidance for media sanitization, including how organizations should select sanitization methods based on the sensitivity of the information stored on the media.

     3. Assets Are Evaluated for Remaining Value

Retirement does not always mean an asset has reached the end of its useful life.

Teams should assess its working condition, remaining useful life, hardware generation, specifications, secondary-market demand, repair needs, testing costs, and removal or transportation costs.

This assessment helps determine if the asset should be redeployed, resold, used for parts, recycled, or disposed of.

Where Do Retired Data Center Assets Go?

Retired data center assets can follow different paths depending on their condition, remaining useful life, data security requirements, and market value.

Below are the main routes these assets can take after retirement.

     1. Internal Redeployment

Usable equipment does not always need to leave the organization. Servers, storage systems, networking equipment, and other hardware may be moved to another data center, a backup environment, a development or testing environment, another facility, or a lower-priority workload.

Internal redeployment can extend the useful life of equipment that no longer meets the requirements of its original environment. It can also reduce the need to purchase replacement hardware elsewhere in the business.

“Companies retire equipment from one data center even when another location still needs the same type of asset. When teams have visibility into available equipment across the business, internal redeployment can save more than 85% compared with buying new,” said Luke Crihfield, Senior Director of Revenue at Amplio, which manages surplus and retired equipment for enterprises through its surplus asset management services. “The key is identifying that internal demand early, before usable equipment moves into resale, recycling, or another disposition route.”

     2. Resale and Remarketing

Assets that no longer serve an internal need may still have value in the secondary market.

Servers, storage equipment, network hardware, memory, CPUs, racks, and power equipment may attract buyers based on their condition, age, hardware generation, and specifications.

Timing also matters. As newer generations become available, demand for older hardware can decline, which can reduce resale value.

     3. OEM Return or Trade-In

Some retired equipment may qualify for an original equipment manufacturer (OEM) takeback, trade-in credit, lease return, or vendor buyback option.

These routes can provide a structured way to remove equipment and recover some value. Companies should compare the expected return with internal reuse and secondary-market resale before choosing this route.

     5. Parts Harvesting

An asset with limited value as a complete unit may still contain components that can be reused or sold.

CPUs, RAM, power supplies, controllers, fans, network cards, and storage components may retain value even when the full server or system has little market demand. These parts can support existing equipment, serve as replacement inventory, or be sold through secondary markets.

Parts harvesting can recover value that would otherwise be lost if the entire asset were sent directly to recycling.

     6. Recycling and Final Disposal

Equipment with no practical reuse, resale, or parts recovery option can move into electronics recycling.

Recyclers can recover materials such as copper, aluminum, steel, plastics, and circuit boards from retired hardware and supporting infrastructure. This route captures material value from equipment that has reached the end of its practical service life.

Final disposal should come only after reuse, resale, parts recovery, and recycling options have been assessed.

Why ITAD Does Not Cover the Entire Refresh

IT asset disposition, or ITAD, typically focuses on electronic and data-bearing equipment such as servers, storage devices, drives, and networking hardware.

A broader data center refresh can leave behind other assets, including racks, UPS systems, batteries, cooling equipment, power infrastructure, electrical components, and other facility equipment.

Many of these assets sit outside the core scope of traditional ITAD, but they still need a clear disposition route. Depending on their condition and remaining value, they may need to be evaluated for reuse, resale, recycling, or disposal.

This is why a full data center refresh often requires a broader asset disposition and recovery process alongside ITAD.

Conclusion

A data center refresh does not end when the new technology goes live. Retired equipment still needs to be secured, assessed, and assigned the right next step.

The right disposition route depends on the asset’s remaining operational, financial, and material value. Reuse, redeployment, resale, and parts recovery should be considered before equipment moves to recycling or final disposal.

A planned disposition process helps protect data, recover more value, and ensure retired data center assets move to the most appropriate next use.

# # #

About the Author

Luke Crihfield is a senior director of revenue at Amplio, helping manufacturers turn surplus into opportunity through AI-driven growth.

The post What Happens to Data Center Assets After a Technology Refresh? appeared first on Data Center POST.

TL;DR

• Multiple fuel vendors do not necessarily create resiliency. Without one partner accountable for the full path from fuel source to generator, handoffs between vendors can create additional gaps, delays, and points of failure.
• Having fuel onsite is not the same as being fuel-ready. True readiness depends on fuel quality, inventory, storage and transfer equipment, emergency supply, and the ability to reliably deliver usable fuel to the generator when needed.
• End-to-end fuel lifecycle management strengthens accountability and operational readiness. A single partner coordinating fuel quality, monitoring, maintenance, delivery, and emergency response can reduce complexity and help ensure the entire system performs as intended.

# # #

When a facility’s backup power depends on diesel fuel, the biggest risk isn’t always a generator failure; it’s the gaps between vendors. Fuel procurement, storage, filtration, testing, emergency delivery, and equipment servicing often sit under separate contracts, each with its own priorities and response times. In a crisis, that fragmentation can mean delayed deliveries, unclear accountability, and ultimately, downtime.

Just as importantly, having fuel onsite is not the same as being fuel ready. A full tank does not guarantee that the fuel is usable, that the transfer system will perform as intended, or that additional supply can be delivered when an outage extends beyond initial runtime assumptions. True readiness depends on the entire fuel pathway working together.

The Hidden Costs of Vendor Fragmentation

Many organizations assume that using multiple specialized fuel vendors creates redundancy and reduces risk. In practice, multiple vendors do not necessarily equal resiliency. A fuel supplier may be responsible for delivery but not stored-fuel quality. A filtration provider may maintain fuel condition but have no visibility into inventory or transfer-system performance. A testing lab may identify a problem without having responsibility for corrective action or emergency logistics.

Without one party accountable for the complete path from fuel source to generator, each handoff can become an additional point of failure. The issue is not simply how many vendors are involved, but whether any one is responsible for ensuring that the entire system performs as it should.

When something goes wrong – say, water contamination is found during routine testing – each vendor can point to another part of the chain. The result is slower resolution, higher costs from emergency orders, and increased operational risk. Industry guides on supply chain resilience emphasize that clear accountability and coordinated procedures are essential for maintaining fuel readiness, especially during disruptions.

The alternative is a model where one partner manages  the entire fuel lifecycle, from terminal to engine.

What End-to-End Management Looks Like

End-to-end fuel lifecycle management means one provides accountability across the entire journey of fuel: from selection at the terminal, through storage and quality maintenance, to delivery and engine readiness. This model aligns incentives: The same team that designs the filtration system also monitors inventory, conducts testing, and stands ready to deliver emergency fuel.

Key elements include:

• Unified accountability: A single point of contact for all fuel-related issues, reducing finger-pointing and speeding up decisions.
• Integrated monitoring: Continuous tracking of fuel quality, inventory levels, and equipment performance to catch issues before they escalate.
• Coordinated emergency response: Pre-planned logistics that ensure fuel and equipment support arrive together when needed most.

Why It Matters Now

Recent supply chain disruptions have highlighted how fragile fragmented systems can be. Rising fuel costs, transportation bottlenecks, and stricter environmental regulations add pressure to get fuel management right. ESI Total Fuel Management approaches fuel as a complete operational system, connecting supply, stored-fuel quality, inventory, infrastructure, maintenance, emergency logistics, and generator readiness under one accountable model.

For mission-critical facilities like data centers, hospitals, and emergency operations centers, the stakes are especially high. A generator that fails to start due to fuel contamination or delayed delivery can have cascading effects far beyond the facility itself.

Practical Takeaways for Facility Leaders

• Map your fuel lifecycle: Industry resilience guides, including FEMA’s supply chain guidance, recommend mapping every step from procurement to engine readiness to identify accountability gaps.
• Consolidate where possible: Consider whether a single partner can manage more of the lifecycle to reduce coordination overhead.
• Test for readiness, not just compliance: Routine testing should validate that fuel will perform under real emergency conditions, not just meet baseline standards.
• Plan for the worst-case scenario: Ensure your fuel partner can deliver both fuel and equipment support simultaneously during an outage.

A Clearer Path to Fuel Readiness

The goal of fuel management isn’t just to keep tanks full; it’s to ensure that backup power systems start reliably when needed most. End-to-end lifecycle management  reduces complexity, clarifies accountability, and aligns every part of the fuel system toward that goal. For organizations that depend on uninterrupted power, having one partner responsible for the entire fuel lifecycle isn’t just convenient – it’s a strategic advantage.

To discuss what a dedicated fuel partner can mean to your operation, contact ESI Total Fuel Management today.

The post End-to-End Fuel Lifecycle Management: Why One Partner Beats Multiple Vendors appeared first on Data Center POST.

TL;DR

• The global data center immersion cooling market is projected to expand from USD 1.7 billion in 2025 to USD 10.9 billion by 2035 at a 19.8% CAGR, propelled by accelerating artificial intelligence and machine learning workloads, rising rack power densities, and mounting energy efficiency pressures.
• Immersion cooling delivers significantly higher compute density per square foot compared to conventional air-cooled systems, allowing colocation and urban data center operators to maximize existing infrastructure without major capital expansion.
•  The solution segment led the component market with a 72% share (USD 1.2 billion) in 2025, while hyperscale operators held a 29.8% share by end user as they prioritize total cost of ownership, operational efficiency, and long-term sustainability.

# # #

The global data center immersion cooling market was valued at USD 1.7 billion in 2025 and is estimated to grow at a CAGR of 19.8% to reach USD 10.9 billion by 2035, according to the recent report by Global Market Insights Inc.

Market growth is entering a high-growth phase, driven by accelerating digital transformation and next-generation computing requirements. Market expansion is fueled by the rapid surge in artificial intelligence and machine learning workloads, increasing rack power densities, and mounting pressure to improve energy efficiency across facilities. As computing intensity continues to escalate, traditional cooling approaches are becoming less viable, accelerating the shift toward immersion-based systems. Rising electricity costs and heightened environmental accountability are further reinforcing the economic case for more efficient thermal management technologies.

Limited floor space in colocation environments and urban data facilities is also encouraging operators to adopt high-density solutions. Immersion cooling enables significantly higher compute density per square foot compared to conventional air-cooled systems, allowing facility operators to maximize existing infrastructure without major capital expansion. In single-phase immersion cooling architectures, dielectric fluid remains in a liquid state throughout operation, while pumps and external heat exchangers effectively dissipate accumulated thermal loads.

By component, the data center immersion cooling market from solution segment held a 72% share, generating USD 1.2 billion in 2025. This segment encompasses integrated immersion cooling systems, including immersion tanks, coolant management frameworks, thermal exchange units, monitoring platforms, and supporting integration hardware. Vendors are increasingly delivering comprehensive, ready-to-deploy systems designed for seamless compatibility with established data center environments. Modern solutions incorporate intelligent temperature tracking, automated coolant regulation, and connectivity with facility management systems to enhance operational performance and overall energy optimization.

In terms of end users, the hyperscale segment held 29.8% share in 2025 and is forecast to reach USD 3.5 billion by 2035. Large-scale data center operators are heavily investing in advanced computing infrastructure to support high-performance processing demands. These facilities operate extensive server environments within purpose-built spaces, prioritizing efficiency, scalability, and long-term infrastructure resilience. Hyperscale operators place strong emphasis on total cost of ownership, carefully balancing capital expenditures, operating efficiency, and long-term sustainability objectives when selecting cooling technologies.

U.S. Data Center Immersion Cooling Market generated USD 454.2 million in 2025 and is anticipated to grow at a CAGR of 16.9% from 2026 to 2035. Increasing demand for advanced cooling technologies is driven by the expansion of large cloud service providers and major technology enterprises headquartered in the country. As digital infrastructure footprints expand, organizations are prioritizing cost-effective and energy-efficient thermal management systems to maintain operational competitiveness. Strong regional commitments to sustainability and improved power utilization effectiveness are accelerating the adoption of immersion cooling solutions across the U.S. market.

Key participants operating in the global data center immersion cooling market include Vertiv, Dell Technologies, Fujitsu, Submer, Green Revolution Cooling, LiquidCool Solutions, Asperitas, Bitfury, DCX Liquid Cooling Company, and Midas Immersion Cooling. Companies in the market are reinforcing their competitive positioning through continuous product innovation, strategic collaborations, and geographic expansion. Industry players are investing in research and development to enhance system reliability, thermal efficiency, and scalability to meet evolving high-density computing requirements. Many firms are introducing modular and turnkey solutions to simplify deployment and accelerate time to market. Strategic alliances with data center operators, colocation providers, and hyperscale customers are strengthening long-term supply agreements and expanding global reach. Businesses are also focusing on interoperability with existing infrastructure to reduce adoption barriers.

The post Data Center Immersion Cooling Market to hit USD 10.9 billion by 2035 appeared first on Data Center POST.

TL;DR

• As high-density GPU clusters and liquid cooling reduce traditional operating margins, procedural discipline and human error are becoming increasingly important factors in AI infrastructure resilience.
• Operational Complexity of Liquid Cooling: Direct-to-chip and immersion liquid cooling solve thermal barriers but introduce complex operational demands, such as coolant chemistry, leak detection, and CDU coordination, where minor procedural mistakes carry immediate, expensive consequences.
• Focus on Long-Term Operational Maturity: AI infrastructure, especially in emerging markets like Central Asia, will ultimately be judged by long-term operational uptime and disciplined management rather than initial construction speed

# # #

As GPU clusters get denser and liquid cooling becomes standard for high-end AI, the decisive variable in AI infrastructure is shifting from how a facility is built to how it is run. Operational standards, not just megawatts, are becoming a critical measure of readiness.

Every conversation about AI infrastructure opens with the same numbers: megawatts secured, racks pushing past 100 kW, GPU counts, gigawatt campuses, interconnect queues. Those numbers matter without power and somewhere to put the heat, nothing scales. But they describe what a facility can theoretically hold, not whether it can carry that load safely at 3 a.m., when a coolant distribution unit alarms and the on-call technician has minutes to decide whether to fail over, shed load, or ride it out. That decision isn’t made by a GPU or a chiller. It’s made by a person following or failing to follow a procedure. And that is where AI data centers are now most exposed.

AI changes the operating model, not just the hardware

Traditional facilities were forgiving: predictable workloads, mature air cooling, settled procedures. AI breaks that. OCP describes AI-driven infrastructure as an order-of-magnitude increase in rack power density, with roadmaps pointing towards 1 MW racks in the next few years. Power volatility adds another challenge: NVIDIA has described rack-level AI workloads swinging from around 30% to 100% utilization and back again in milliseconds, reflecting the synchronized compute and communication phases documented in large-scale AI training research. Loss of coolant flow can quickly translate into thermal throttling, instability or protective shutdowns. At these densities, the coupling between IT load, power and cooling becomes tight and unforgiving. The hard problem is no longer only construction; it’s operational maturity, and the gap between the two is where expensive clusters get lost.

Liquid cooling solves a thermal problem and creates an operational one

Nowhere is this clearer than with liquid. Direct-to-chip liquid cooling is becoming increasingly standard for high-density AI, while immersion remains one of several approaches, but moving water or dielectric fluid through a live data hall introduces operational disciplines that may be unfamiliar to teams accustomed to air-cooled environments: coolant chemistry and conductivity, biological growth and corrosion in the cooling loop, filtration, leak detection and containment, and constant coordination between the facility water system and the rack-level CDU. The OCP Project Deschutes CDU specification targets roughly 2 MW of cooling capacity at 500 GPM with 80 to 90 psi of available pressure, with defined coolant and system requirements for high-density liquid-cooled deployments. At this scale, liquid cooling becomes an operational system in its own right, with consequences for maintenance, monitoring and response procedures. It does not remove the operational problem. It raises the stakes on it.

Human error is now an infrastructure risk

Uptime Institute’s 2025 outage analysis found that 85% of major human-error outages involved either staff failing to follow procedures or flaws in the procedures themselves, and that nearly 40% of organizations had suffered a major outage caused by human error over the previous three years. Its 2026 analysis shows that failure to follow established procedures remains the leading driver of human-error-related outages. Power is still the leading cause of impactful outages overall, so the point is not that procedures replace infrastructure risk. It is that procedures remain one of the clearest controllable weaknesses once the facility is live: whether the runbook exists, whether it is correct, whether anyone follows it, and whether change management catches the conflict before it reaches the floor. In lower-density environments, operational mistakes often had more thermal and electrical margin around them. At 100 kW per rack and above, particularly in liquid-cooled environments, those margins can shrink sharply. The same procedural error can therefore have much faster and more expensive consequences.

What “standards-based operations” actually means

The phrase has to stop being a slogan. It means a facility runs on documented, auditable practice rather than tribal knowledge and real frameworks exist for this. The operations layer has its own independent assessment framework: Uptime Institute’s Management & Operations (M&O) Stamp of Approval evaluates staffing, organization, training, maintenance, operating conditions, planning and coordination independently of the Tier Classification System. A highly resilient design can still be undermined by weak operations. Around it sits the management-system scaffolding the best operators already use: ISO/IEC TS 22237-7:2018, which specifically addresses data centre management and operational processes, ANSI/TIA-942-C for physical infrastructure, and ITIL-style change and incident management that turns change control from a form into a real risk gate. In practice that means clear methods of procedure and emergency operating procedures; change management that accounts for how one modification now ripples across load, cooling, power, and network at once; staff trained for liquid and high density rather than retrofitted from air; telemetry that correlates IT load with thermal and power behavior in real time; and risk management that prices in operational error.

The industry is already moving this way

The clearest signal is OCP’s Open Data Center for AI initiative, launched in October 2025 alongside an industry call for collaboration initiated by Google, Meta and Microsoft and expanded through new workstreams and specifications in 2026. It is explicitly about standardizing not only power and cooling hardware but management telemetry and operational technology and it names the operational problem directly: AI facilities now involve several parties at once (facility owner, hardware owner, workload owner), making service-level agreements and coordination far more complex than the clean colocation SLAs of the air-cooled era. When the largest operators on the planet start writing shared operating specifications, operations not just construction has become a competitive frontier.

Emerging markets: the sharpest version of the problem

This is where I watch it play out most acutely across Central Asia, where AI capacity, power supply, an engineering labor market, and an operating culture are all being built at once. The risk here isn’t only a shortage of megawatts or GPUs; it’s a maturity gap in procedures, people, and change management that never appears on a capex spreadsheet and surfaces only after the first avoidable incident. But the asymmetry cuts the other way too: a greenfield operator carrying no legacy habits can adopt M&O-grade procedures, real change management, and proper telemetry from day one, instead of bolting them on after a decade of “this is how we’ve always done it.” Framed correctly, standards-based operations aren’t bureaucracy that slows an emerging market down. They are the fastest route to closing the reliability gap with mature regions and to skipping mistakes the industry has already paid for.

The real test comes after the ribbon-cutting

AI data centers are becoming national digital infrastructure, and they will be judged not on how fast they broke ground but on whether they run. Power makes deployment possible, cooling keeps density stable, GPUs do the compute but the facility that earns trust over years is the one with correct procedures, trained people, disciplined telemetry, real change control, and risk management that treats human error as seriously as equipment failure. The strongest AI data centers won’t be the ones built first. They’ll be the ones still running predictably while everyone else is firefighting.

# # #

About the Author

Maksim Gavriliuk is an independent digital infrastructure and data center expert, IEEE Senior Member, and DCOS 2026 Review Committee Member. He specializes in data center operations and AI-ready infrastructure, with over 20 years of experience in IT infrastructure projects across Central Asia and Eastern Europe.

The post AI Data Centers Don’t Fail at the Rack. They Fail in the Runbook. appeared first on Data Center POST.

TL;DR

• Floating data center site selection is a system design decision rather than a simple real estate exercise, requiring location, engineering, and regulation to be managed in parallel to solve land, power, and cooling constraints.
• Developers must evaluate six core factors: energy availability and scalability, marine environment metocean conditions, operational logistics, digital connectivity redundancy, regulatory permitting pathways, and physical/cyber security.
• Long-term viability requires managing key operational trade-offs over multi-decade lifespans, balancing near-shore accessibility against offshore expansion potential and grid reliance versus dedicated offshore generation.
• Early integration of marine risk assessment alongside classification societies like ABS avoids costly late-stage redesigns and provides clarity across complex regulatory jurisdictions

# # #

When it comes to floating data centers, it really is about ‘location, location, location’. Site selection influences engineering, regulation, operations and long-term resilience, which is why it pays to get this right from the outset, writes Frank Genco, ABS Vice President of Strategic Account Program 

When it comes to floating data centers (FDC), the question isn’t why offshore. It is increasingly where. The whys are clear: the rapid growth of data centers is already testing grid capacity, water resources and the tolerance of local communities living alongside these facilities. By siting data centers offshore, however, many of those constraints can be resolved, with developers able to access new energy sources, abundant seawater for cooling and locations unconstrained by land availability.

But where? That’s a bigger question, and the answer will influence the regulatory pathway, engineering design, reliability, cost, maintenance, security and uptime for the next 30 years.

More than a real estate decision

This isn’t a site selection checkbox exercise, however. It’s a system design decision, because location, engineering and regulation are closely interconnected. A change to one often affects the other two. A site that appears attractive from an operational perspective may introduce regulatory complexity. Engineering innovations may suddenly make previously unsuitable locations viable. These three strands (location, regulation and engineering design) need to be managed in parallel because it is at their intersection where the right solution in the right location will be found.

 ABS brings decades of experience supporting complex marine and offshore assets, from floating LNG and ultra-deepwater oil production units to floating wind. That experience can help engineers understand how location choices influence requirements and regulatory pathways long before detailed design begins.

Every answer creates another question

Site selection involves identifying and testing a wide range of competing and interdependent issues all at the same time. An apparently straightforward question about available infrastructure, such as natural gas pipeline proximity, may lead to broader engineering considerations. Does a location provide sufficient access to electrical power today for a standalone FDC, and can it support future expansion or augmentation? Would dedicated offshore generation offer greater resilience and public approval than relying solely on the grid? Could nearby renewable energy or future small modular nuclear reactors reshape the long-term energy strategy?

It’s the same for the cooling requirements of the computing and power generation facilities. Engineers need to understand seasonal temperature variation, water movement, corrosion rates and whether engineering trade-offs could expand the range of viable operating conditions. Improving cooling system design may significantly increase the number of potential locations but how would this innovation impact timelines?

Regulation presents another layer of complexity. The applicable framework depends on numerous factors, including proximity to shore, whether the facility is permanently moored or mobile, the pipeline requirements, the presence of subsea power or communications infrastructure and which authorities have jurisdiction. Depending on the location, developers may need to engage with federal, state, local and port authorities, alongside maritime regulators and environmental agencies.

Understanding these interactions early can avoid costly redesigns later in the project.

The major considerations

Every project will have unique requirements, but several key factors will consistently shape site selection. These include:

1. Energy. Developers must determine where power will come from, whether existing grid infrastructure is sufficient, or whether dedicated offshore generation offers greater resilience and flexibility. Equally important is understanding how future demand growth can be accommodated without major redesign.
2. Marine environment. Long-term metocean conditions, extreme weather exposure, wave climate, currents, corrosion potential and marine growth all influence structural design, maintenance schedules and operating costs.
3. Operational logistics. How easily can personnel access the facility? Which nearby ports can support maintenance activities? How resilient are local supply chains? These questions directly affect availability and lifetime operating costs.
4. Digital connectivity. Fiber availability, cable redundancy and network latency all influence performance. Building resilience into communications infrastructure from the outset can significantly reduce operational risk over the facility’s lifetime.
5. Regulatory landscape. Understanding permitting pathways, environmental approval processes, certification requirements and jurisdictional responsibilities early in the project can help avoid delays and provide greater certainty during development. ABS works with developers at the regulatory interface, helping to map the applicable frameworks and identify the right agencies before those conversations become constraints on engineering or schedule.
6. Security. Physical protection, cybersecurity, vessel traffic management, supply chain resilience and geopolitical risk all become increasingly important as critical digital infrastructure moves offshore.

Thinking beyond first deployment

A floating data center will likely have an operational lifespan measured in decades, making resilience and future-proofing central to site selection. The maritime and offshore industries are used to thinking long-term and building offshore structures designed to withstand rare and extreme events, but these floating data factories will be packed with cutting-edge computing technology and serving markets that are evolving at speed, which may require additional forethought. Developers should consider whether today’s optimal location will still be viable if computing demand doubles, cooling technologies evolve or additional modules are needed.

Managing trade-offs

There is rarely a perfect location. Sheltered near-shore waters may simplify operations initially but limit future expansion. Offshore sites may provide greater access to renewable energy but introduce more demanding environmental conditions that require additional engineering and impact costs. Grid-connected facilities may reduce upfront investment but dedicated offshore generation could offer greater resilience and scalability.

These trade-offs are not unique to floating data centers. Offshore developers are used to navigating competing priorities and understand that the key is to evaluate these trade-offs systematically and simultaneously so that technical, commercial and regulatory implications are considered together rather than in isolation. In this way, a holistic, multi-disciplinary and multi-agency approach can be taken that minimizes the risks of unwelcome surprises further down the road.

Involving marine risk assessment from the beginning

Ultimately, site selection is a risk management exercise. The maritime and offshore industries have spent decades developing sophisticated approaches to evaluating operational, environmental and engineering risk. Many of the challenges floating data centers will face have already been addressed across offshore energy, shipping and marine infrastructure. Applying these proven methodologies during the earliest stages of development enables better-informed decisions before major investments are committed. ABS works across all three of these dimensions simultaneously, combining marine risk assessment, regulatory engagement and engineering insight to help developers match the right asset to the right location.

As floating data centers move from concept to commercial reality, the most successful will be those that view site selection as integral to engineering and commercial outcomes, not simply as a location for the “real” work to take place. By working with ABS to integrate engineering, regulation and marine risk assessment from the outset, developers can identify locations that work today and are also resilient, scalable and commercially viable for decades to come.

 # # #

About the Author

Frank Genco is the Vice President of Strategic Accounts at the American Bureau of Shipping (ABS). He created and heads the Strategic Account Management (SAM) Program. His division is tasked with aligning ABS’s global enterprise resources across maritime, energy, digital infrastructure, cyber security, and risk management to service their highest-value global clients. Frank specializes in critical infrastructure, strategic corporate partnerships, and connecting top-tier technology companies with maritime and digital risk expertise. He is currently building out specialized teams, including expanding executive relationships with Silicon Valley technology firms to bridge the gap between AI/digital infrastructure and maritime compliance.

The post Site Selection: The Most Important Decision You’ll Make Before Construction Begins appeared first on Data Center POST.

TL;DR

• Hose selection must be driven by complete system operating conditions (including specific coolant chemistry, thermal cycling, working pressure, permeation, and cleanliness requirements) rather than choosing a material name directly from a catalog page.
• Hose material performance impacts long-term reliability, influencing key operational factors beyond flexibility, such as pressure drops, gradual fluid permeation, and loop contamination.
• Common hose materials involve distinct design trade-offs, with options like EPDM, silicone, PTFE, PFA, and metal offering different balances of chemical compatibility, flexibility, mechanical strength, cost, and permeation resistance.
• Hoses and fittings should be integrated early into overall rack, manifold, and cooling system architecture rather than left as an afterthought during final installation

# # #

By Xiang Wu, IT & Automation Director, CJan Fluid Technology Co., Ltd.

As liquid cooling moves from specialized deployments into mainstream AI and high-performance computing infrastructure, more attention is being paid to pumps, cold plates, cooling distribution units and control systems.

The hose is easier to overlook.

Yet in a direct-to-chip cooling loop, the hose is part of the fluid path that operates continuously while carrying coolant between components that may need to be serviced independently. Its material affects more than flexibility. It can influence pressure capability, chemical compatibility, permeation, cleanliness, bending behavior and service life.

For that reason, choosing a liquid cooling hose should start with the operating conditions of the cooling loop, not with a material name on a catalog page.

Start With the Coolant, Not the Hose

The first question in hose selection should be simple:

What fluid will actually be inside the hose?

Water, water-glycol mixtures and specially formulated cooling fluids do not place exactly the same demands on hose materials. Even when two coolants appear similar, differences in additives, inhibitors, concentration and operating temperature can affect material performance.

This is particularly important for systems that are expected to operate continuously for years.

A material that appears compatible during a short laboratory test may behave differently after thousands of hours of exposure. Swelling, hardening, extraction of low-molecular-weight components and changes in mechanical properties may develop gradually.

For this reason, coolant compatibility should be evaluated using the actual fluid formulation whenever possible.

Material selection should consider:

• Coolant chemistry
• Glycol concentration, where applicable
• Corrosion inhibitors
• Biocides and other additives
• Operating temperature
• Expected service life
• Cleanliness requirements

The hose should be qualified as part of the actual cooling system rather than treated as an isolated component.

Temperature Is More Than a Maximum Rating

Hose specifications often list a maximum temperature. That number is useful, but it does not tell the whole story.

A cooling loop may experience a relatively stable operating temperature most of the time while still going through repeated thermal cycles. Startup, shutdown, maintenance and changes in computing load can all introduce temperature changes.

Repeated thermal cycling can affect:

• Elastomer flexibility
• Reinforcement
• Fitting connections
• Sealing performance
• Long-term material stability

A hose material therefore needs to be evaluated against the complete temperature profile rather than a single maximum value.

For example, EPDM is widely used in water-based cooling applications because of its resistance to water and glycol-based fluids. Silicone offers excellent flexibility and broad temperature capability in many applications. Fluoropolymer materials such as PTFE and PFA can be considered when chemical resistance, cleanliness or low extractables are particularly important.

The point is not that one material is universally better than another.

The right material depends on what the cooling loop requires.

Pressure and Flow Still Matter

Liquid cooling systems are designed around controlled flow, but that does not mean hose pressure requirements can be ignored.

The hose should be evaluated against:

• Normal operating pressure
• Maximum working pressure
• Pressure fluctuations
• Pump characteristics
• Startup and shutdown conditions
• Potential pressure spikes

Internal diameter also matters.

A hose that is too small can increase pressure drop and require additional pump head. A hose that is unnecessarily large may create packaging and routing problems.

The engineering objective is therefore to balance flow requirements with pressure capability, flexibility and available space.

The hose should also be evaluated as a complete assembly. The pressure rating of the tube alone does not necessarily represent the pressure capability of the hose, fitting and connection together.

Flexibility Has a Practical Value

One reason flexible hose continues to have a role in liquid-cooled data centers is serviceability.

Servers, cold plates, manifolds and rack-level cooling components may need to be removed or repositioned during maintenance. A rigid connection can make this difficult.

But flexibility should not be confused with unlimited bending.

Every hose has a minimum bend radius. Repeatedly forcing a hose tighter than its recommended radius can place unnecessary stress on the tube, reinforcement and fittings.

Good routing should avoid:

• Sharp bends near fittings
• Twisting along the hose axis
• Continuous tensile loading
• Contact with sharp edges
• Uncontrolled movement caused by vibration

A slightly longer hose with proper routing can sometimes be more reliable than a shorter hose installed under constant mechanical stress.

This is one reason hose routing should be considered during rack and manifold design rather than left to the final installation stage.

Permeation Deserves More Attention

Leakage is an obvious concern in a liquid cooling system. Permeation is less visible.

Some hose materials allow molecules to gradually pass through the hose wall. The rate depends on the material, fluid, temperature, pressure and wall construction.

For a short-duration application, this may have little practical significance. For a cooling loop expected to operate continuously over a long period, it can become part of the system reliability calculation.

Potential consequences include:

• Gradual coolant loss
• Changes in coolant concentration
• Moisture or gas transfer
• Changes in fluid condition
• Increased maintenance requirements

This does not mean that a low-permeation material is automatically the correct choice.

It means that permeation should be considered whenever coolant stability and long service intervals are important design objectives.

Cleanliness Is a System Requirement

Data center liquid cooling is not simply a plumbing application.

The coolant passes through components such as cold plates, heat exchangers, pumps and manifolds. Contamination introduced by one component can potentially travel throughout the loop.

For systems with tighter cleanliness requirements, engineers may need to consider:

• Extractables
• Particles
• Ionic contamination
• Internal surface condition
• Manufacturing cleanliness
• Cleaning and flushing procedures

This is one area where material selection and manufacturing quality are closely connected.

A material with appropriate chemical resistance may still be unsuitable if the finished hose assembly introduces unacceptable contaminants into the cooling loop.

The Fitting Is Part of the Decision

Selecting the hose and selecting the fitting should not be treated as two independent decisions.

The connection needs to accommodate the hose construction, pressure, temperature, installation method and expected movement.

Particular attention should be paid to the area immediately behind the fitting.

This is often where bending, vibration and installation forces concentrate. If the hose is forced into a sharp bend immediately after the connection, the fitting may effectively become the mechanical anchor point for the entire assembly.

A good installation distributes mechanical loads instead of transferring them directly to the connection.

A Practical Material Selection Approach

There is no single hose material that should be specified for every liquid cooling system.

A practical selection process can begin with six questions:

1. What is the coolant?
   Identify the actual fluid and additives.

2. What are the temperature conditions?
   Consider both steady-state temperature and thermal cycling.

3. What pressure and flow are required?
   Check working pressure, pressure fluctuations, internal diameter and pressure drop.

4. How much movement is expected?
   Determine whether the hose is static, occasionally moved or continuously flexed.

5. What cleanliness and permeation requirements apply?
   These factors may influence the choice between elastomeric and fluoropolymer constructions.

6. How will the hose be installed and maintained?
   Consider bend radius, fitting accessibility, replacement and service clearance.

Only after these questions are answered should the material be selected.

A Simple Comparison

This table should be treated as a starting point rather than a specification guide. Actual material performance depends on the formulation, hose construction, manufacturing process and operating environment.

The Best Hose Is the One That Fits the Whole System

Liquid cooling projects sometimes approach hose selection from opposite directions.

One team may prioritize flexibility.

Another may prioritize pressure.

A third may focus on chemical compatibility or cleanliness.

All of these considerations are legitimate. The challenge is that optimizing one characteristic without considering the others can create a different reliability problem.

A highly flexible hose may not provide the desired pressure or permeation performance. A chemically resistant material may require a different reinforcement or connection method. A mechanically robust assembly may be harder to route in a dense rack environment.

The selection process therefore needs to look at the complete operating envelope.

For high-density AI infrastructure, this becomes increasingly important as liquid cooling systems become more complex and maintenance access becomes more valuable.

Design the Hose Into the Cooling System

The hose should not be the last component selected after the rest of the cooling architecture has already been fixed.

It should be considered alongside the cold plate, manifold, CDU, pump, quick-disconnects and rack layout.

When material, construction, routing and connections are evaluated together, engineers can make better trade-offs between reliability, serviceability and cost.

The objective is not simply to find a hose that can carry coolant.

The objective is to build a fluid path that remains predictable throughout the operating life of the data center.

As direct-to-chip cooling continues to expand, that distinction will become increasingly important.

# # #

About the Author

Xiang Wu is the IT & Automation Director at CJan Fluid Technology Co., Ltd., where he focuses on industrial automation, digitalization, and engineering applications for advanced fluid systems. His work brings together automation technology and practical engineering requirements across demanding industrial applications, including data center liquid cooling and semiconductor equipment. Xiang has a particular interest in how material selection, system design, and component reliability affect the performance of fluid-handling systems. He works closely with engineering teams and customers to develop practical solutions for complex cooling and fluid-transfer challenges.

The post How to Select Hose Materials for Direct-to-Chip Liquid Cooling Systems appeared first on Data Center POST.

TL;DR

• Datalec Precision Installations (DPI) has welcomed 12 new apprentices as part of its second apprenticeship cohort.
• The programme combines hands-on workplace experience with industry-recognised qualifications and structured career development.
• Apprentices will gain experience across technical and operational disciplines supporting mission-critical environments throughout the UK.
• DPI’s apprenticeship programme supports the development of skilled talent as demand for data centre and critical infrastructure expertise continues to grow.
• Edmundson Electrical and Safety Signs are supporting the new cohort with professional tools, PPE and uniforms.

###

As investment in cloud computing, digital services and AI continues to drive data centre growth, the need for skilled professionals capable of supporting critical digital infrastructure is growing alongside it.

Datalec Precision Installations (DPI) is helping build that talent pipeline by welcoming 12 new apprentices to its workforce as part of the second cohort of its apprenticeship programme. The initiative provides young people with a direct pathway into the data centre industry through a combination of hands-on experience, formal training and structured career development.

The new apprentices will begin their careers across a range of technical and operational disciplines, working alongside experienced DPI professionals on projects supporting mission-critical environments throughout the UK.

As data centres play an increasingly important role in the UK’s digital economy, developing the workforce needed to build and support that infrastructure has become an industry-wide priority. DPI’s apprenticeship programme is designed to help address that skills challenge while creating meaningful, long-term career opportunities for young people.

“Apprenticeships provide an outstanding route into our industry, enabling individuals to earn while they learn and develop practical skills within a real working environment,” said Danny Turley, Learning & Development Manager at DPI. “As demand for digital infrastructure continues to grow, investing in future talent has never been more important.”

Through the programme, apprentices receive mentorship, structured on-the-job learning and exposure to a wide range of disciplines across the business. This approach provides participants with technical expertise, professional qualifications and transferable skills that can form the foundation for long-term careers within the sector.

DPI’s industry partners are also supporting this year’s cohort from day one. Edmundson Electrical has supplied professional tools for the apprentices, while Safety Signs has provided PPE and uniforms.

With continued investment in emerging talent, DPI is helping develop the next generation of skilled professionals equipped to build, install and support the critical digital infrastructure underpinning modern business and society.

The programme currently has an apprenticeship retention/completion rate of 100%, with seven business departments participating, demonstrating DPI’s commitment to integrating apprenticeship opportunities across its operations.

Read more in the press release here.

The post Building the Next Generation of Data Centre Talent: Datalec Welcomes New Apprentice Cohort appeared first on Data Center POST.

TL;DR

• ZincFive launched the ZincFive BMS, a next-generation battery management system built specifically for its nickel-zinc (NiZn) chemistry rather than adapted from lithium-ion or lead-acid platforms.
• The system ships standard with every ZincFive BC Series UPS Battery Cabinet and uses a streamlined two-tier architecture (Controller plus Tray Monitor) that simplifies wiring, cuts installation time, and removes the need for third-party network switches.
• In BC 2 AI cabinets, the BMS enables AI pulse intelligence, allowing the system to respond to the high-frequency power swings created by AI training and inference workloads.
• Built-in encryption, authentication, and access controls address growing operational technology (OT) cybersecurity requirements, while a redesigned interface gives operators fleet-wide visibility, remote configuration, and firmware updates across multi-site deployments.

# # #

AI is changing the shape of electricity demand inside the data center, not just the size of it. Instead of a steady draw, GPU clusters can ramp power up and down within milliseconds to seconds, and research from the Electric Power Research Institute (EPRI) has found these swings can reach 40% or more of total load in that window. Backup power systems built for slow, predictable demand were never designed to track that kind of movement, which is why the conversation in the industry has shifted from “how much battery capacity do we have” to “how well can that battery communicate and respond.”

That shift is putting a spotlight on a part of the UPS stack that used to be an afterthought: the battery management system, or BMS. It’s also the reason ZincFive‘s recent launch of a purpose-built BMS is worth a closer look, not just as a product update, but as a signal of where the industry is headed.

The BMS Was Never Meant to Be an Afterthought

A BMS handles the unglamorous work of watching over individual cells and modules, tracking voltage, temperature, and state of charge, and flagging problems before they become failures. For years, many battery suppliers adapted BMS platforms originally built for lithium-ion or lead-acid chemistries and applied them broadly across product lines.

That approach works reasonably well when battery behavior is uniform across chemistries. It works less well when a chemistry has different electrical characteristics, cycle patterns, or failure modes that a borrowed platform was never tuned to interpret. ZincFive’s new BMS takes the opposite approach: It was designed and assembled in the U.S. specifically around nickel-zinc chemistry, using a two-tier hardware architecture, the ZincFive Controller and ZincFive Tray Monitor, built to deliver high-resolution telemetry across an entire battery cabinet rather than retrofitting a platform meant for a different chemistry.

“The ZincFive BMS reflects everything we’ve learned from managing and deploying nickel-zinc batteries at scale,” said Tod Higinbotham, CEO of ZincFive. “By bringing development in-house and assembling the BMS in the U.S., we have greater control over the platform and how it evolves. As AI reshapes data center power requirements, this level of intelligence and control becomes increasingly critical.”

Millisecond Response Requires Millisecond Visibility

Handling AI-driven power transients isn’t only about raw battery chemistry. It also depends on how quickly a system can sense a load change and coordinate a response across every cabinet in a fleet. That’s the specific problem ZincFive’s AI pulse intelligence capability, built into the BMS for BC 2 AI UPS Battery Cabinets, is designed to address: helping the system respond directly to the high-frequency power pulses associated with AI training and inference.

High-resolution, cabinet-level telemetry, tracking individual battery voltage and temperature in near real time, gives operators the visibility needed to confirm that backup power is actually behaving the way it’s supposed to under stress, not just during a scheduled test. Simplifying the physical architecture matters too. ZincFive’s design daisy-chains cabinets via dual Ethernet ports and eliminates the need for a separate data-aggregation layer, an approach aimed at reducing installation time and long-term maintenance burden, a practical benefit that often gets overlooked in conversations focused purely on power density.

Cybersecurity Has Quietly Become a Power Infrastructure Issue

Battery management systems are networked, which means they’re now part of the data center’s operational technology attack surface. Recent industry research has found that a majority of organizations have BMS devices affected by known, exploitable vulnerabilities, a statistic that would have seemed out of place in a power conversation a few years ago. ZincFive built encryption, authentication, hardened communications, and access controls into the new BMS from the ground up, an approach meant to meet increasingly strict OT security expectations while giving operators audit-ready visibility into their systems.

The Takeaway for Data Center Operators

The lesson emerging across the industry is straightforward: battery performance and battery intelligence are no longer separate conversations. ZincFive’s launch, which also adds a redesigned interface for fleet-wide monitoring, remote configuration, and firmware updates across multi-site deployments, is one clear example of that shift in practice. As data centers prepare for an AI-driven power profile that looks nothing like the one they were originally built for, the companies rethinking how batteries are monitored and managed, not just how they’re built, are the ones setting the pace.

To read the full press release, visit here.

The post Why Battery Intelligence Is Becoming as Important as Battery Power appeared first on Data Center POST.

TL;DR:

• A new pan-European data center design and engineering group has entered the market.
• Hyceon, formed from the merger of TTSP HWP and MCA Architects, brings together more than 270 professionals and experience across over 10 GW of data center capacity under one unified organization.
• Its launch reflects a broader shift in how large-scale digital infrastructure projects are being delivered across Europe, at a moment when demand for AI-ready facilities is outpacing traditional project models.

# # #

Frankfurt and Dublin-based Hyceon officially launched this week as the unified entity of two long-established firms, TTSP HWP and MCA Architects. The move completes a merger first announced in December 2025 and marks the arrival of a new pan-European group focused specifically on data center design, engineering, and advisory services.

The timing is notable. Europe’s data center sector is in the middle of a build-out driven by cloud computing and artificial intelligence, and the scale of that demand is reshaping what clients expect from their design and engineering partners. Hyceon’s launch is one of the clearer signals yet that the market is consolidating around firms capable of managing a project from concept through completion, rather than handing it between specialists at each stage.

What the Combined Group Looks Like

Under CEO Alexander Hauser, Hyceon operates from six offices across Germany and Ireland, Frankfurt, Dublin, Stuttgart, Cologne, Berlin, and Düsseldorf, with additional delivery offices in Budapest and Kuala Lumpur supporting projects across the continent. The group brings together more than 270 professionals working across architecture, engineering, master planning, and project management.

Combined, the two founding firms have worked on data center projects representing more than 10 GW of capacity and have carried several hundred schemes through the permitting process. That track record spans five years of prior collaboration between TTSP HWP and MCA Architects, during which their joint data center work already represented over a gigawatt of capacity for hyperscale and colocation clients. The new structure formalizes what had effectively been a working partnership into a single organization.

Why One Point of Accountability Matters Now

Large data center projects typically involve a long sequence of distinct phases: site selection, feasibility studies, permitting, master planning, architecture, engineering, and eventually construction oversight. When each phase is managed by a different firm, coordination gaps can slow projects down or create costly rework later.

“Our clients are building at a speed and scale Europe has not seen before, and they are looking for partners who can carry a whole project rather than a piece of it,” said Hauser. “Hyceon gives them one team and one point of accountability, from site selection to handover.” That structure is particularly relevant as European regulation grows more specific. Germany’s Energy Efficiency Act, for example, introduces tightening PUE and heat-reuse requirements for facilities commissioned from July 2026 onward, standards that are far easier to meet when efficiency planning is built into a project from its earliest stages rather than layered on afterward.

A Signal for What’s Ahead

Hyceon’s launch adds a new, sizable player to a European data center market that continues to attract hyperscalers, cloud providers, colocation operators, and infrastructure investors. As Managing Director Brian Murphy noted, the group intends to expand its footprint alongside its clients: “As our clients expand into new and existing regions, they need partners who can provide the same high standard of service wherever they go.”

For an industry navigating rapid AI-driven growth alongside increasingly detailed regulatory requirements, Hyceon’s formation illustrates a broader pattern worth watching: The firms best positioned to support Europe’s next wave of data center development may be the ones built to manage the entire project, start to finish, rather than a single piece of it.

To read the full launch announcement, please click here. To learn more about what Hyceon brings to the European market, visit hyceon.com.

The post A New Name Enters Europe’s Data Center Market appeared first on Data Center POST.

TL;DR

• Toptana Technologies has been named a winner of the 2026 Broadband Communities Tribal Lands Award.
• The award recognizes Toptana’s work developing new fiber and subsea connectivity infrastructure in Washington State.
• Its network is being built to serve the Quinault Indian Nation while giving the community a direct stake in the infrastructure.
• The project will connect the Washington coast with Seattle, Hillsboro and international markets across the Pacific.

# # #

For the Quinault Indian Nation, building better connectivity starts with a simple question: who gets to decide what that infrastructure looks like and who benefits from it?

Toptana Technologies, the Indigenous-owned telecommunications infrastructure company owned by the Quinault Indian Nation, is taking a different approach. Rather than viewing broadband infrastructure simply as something delivered to a community, Toptana is building and owning infrastructure that can serve the community while creating opportunities beyond it.

That work has earned Toptana the 2026 Broadband Communities Tribal Lands Award, which recognizes efforts to bring modern telecommunications services to tribal lands while respecting sovereignty and addressing gaps in connectivity.

“From the beginning, Toptana has been about creating greater opportunity for the Quinault Indian Nation through connectivity while maintaining our sovereignty and protecting the resources that are central to our people and way of life,” said Tyson Johnston, 1st Council person of the QIN and Ex Officio Board Member. “This recognition is meaningful because it reflects both what we are building and why we are building it to create lasting benefits for our community and future generations.”

The infrastructure behind that effort is already taking shape.

Toptana has completed an approximately 200-mile fiber route running north to south between Seattle and Hillsboro. The company is also advancing an approximately 90-mile east-to-west route that will connect its planned Washington State Cable Landing Station with the I-5 corridor.

The cable landing station is particularly significant for the region. It will be the first new cable landing station developed in Washington State in more than 25 years and is designed to support as many as four subsea cable systems. Once in service, it will provide a new landing point for cables connecting the Pacific Northwest with markets throughout the Asia-Pacific region.

The terrestrial network will carry that international connectivity inland, linking the Washington coast with major network hubs in Seattle and Hillsboro.

“Toptana represents a different approach to developing critical communications infrastructure, one where the community is not simply served by the infrastructure but has ownership and a direct role in its development,” said Joel Ogren, Toptana General Manager and CEO of Assured Communications. “The combination of subsea and terrestrial connectivity creates opportunities not only to improve access locally, but also to establish the region as an important new connection point within the global communications network.”

For the Quinault Indian Nation and surrounding communities, the value of the project will extend well beyond the network itself. Better access to high-speed connectivity can support students and educators, expand telemedicine options, help local companies reach customers and create new opportunities for people working in the region.

Those goals are closely tied to why Toptana was created in the first place. The company is building communications infrastructure with the Quinault Indian Nation involved not just as a customer, but as an owner.

The Broadband Communities recognition puts a national spotlight on that approach. For Toptana, however, the work continues with fiber routes under construction, a new cable landing point taking shape and a long-term objective of giving the Pacific Northwest—and the communities connected to it—a stronger position in the global communications network.

Learn more in the full announcement here.

The post Toptana Technologies Wins 2026 Broadband Communities Tribal Lands Award appeared first on Data Center POST.

TL;DR

• Technical site viability does not necessarily mean a project will have a clear path to community acceptance and approval.
• Social and political intelligence can complement traditional land, power, connectivity and permitting analysis.
• Public records, local news and previous development decisions can provide early indicators of community and political dynamics.
• Earlier risk assessment can help inform site selection, communications, permitting and stakeholder engagement strategies.
• Understanding local concerns before announcing a project can help development teams enter a market better prepared to engage with stakeholders.

# # #

As data center development expands into new markets, land, power, connectivity and permitting are no longer the only factors that can determine whether a project moves forward. Community sentiment, political dynamics, regulatory considerations and local concerns can create significant challenges for projects that otherwise appear viable on paper.

On September 24, iMiller Public Relations (iMPR) and Flypower will host “Assessing Social and Political Risk Before a Data Center Development Goes Public: Know Before You Go,” a live webinar examining how developers, investors and site selection teams can identify and assess these risks earlier in the development process.

Looking Beyond Technical Viability

Traditional site selection provides critical insight into whether a location can support a data center, but it may not fully account for how local stakeholders will respond to the development. Issues involving power and water use, land use, environmental impact, tax incentives, noise and community benefits can quickly become part of the public discussion.

Evaluating the social and political environment earlier adds another layer of due diligence. Municipal and county meeting minutes, voting records, local news coverage and the history of previous development proposals can provide insight into how local leaders and residents have responded to infrastructure projects and which issues could influence a project’s path forward.

Connecting Risk Assessment With Community Engagement

Identifying potential challenges before a project is announced can also help development teams shape communications and stakeholder engagement around the specific dynamics of a market. Instead of reacting to concerns after they emerge, teams can enter the community with a clearer understanding of local priorities and potential areas of opposition.

The September 24 webinar will connect Flypower’s market risk analysis with iMPR’s Community Groundswell program to explore how early intelligence can inform site selection, permitting, messaging and stakeholder strategy. The discussion will also address how developers can identify potential opposition, misinformation and community concerns before they become project-defining issues.

As data center development reaches more communities, understanding where a facility can be built and where it can realistically move through the approval process are becoming increasingly connected. Incorporating social and political intelligence earlier can help project teams make better-informed decisions before significant resources and capital are committed.

Want to learn how to assess these risks before a data center project goes public? Join iMPR and Flypower on Thursday, September 24 at 1 p.m. EDT / 10 a.m. PDT for “Assessing Social and Political Risk Before a Data Center Development Goes Public: Know Before You Go.”

The webinar is free, but advance registration is required. Register today to reserve your spot.

The post Beyond Site Selection: Assessing Social and Political Risk in Data Center Development appeared first on Data Center POST.

TL;DR

• New infrastructure: Toptana Technologies is breaking ground on a new cable landing station on Washington’s Pacific coast.
• First cable: The facility will serve as the Washington landing point for AWS-owned Sta’O’Nuk, the first subsea cable system planned for the site.
• Built for growth: The landing station is designed to accommodate up to four subsea cable systems, creating room for future connectivity.
• Indigenous-led development: The project represents an investment by the Quinault Indian Nation in long-term economic opportunity and Washington’s role in the global digital economy.

# # #

For more than two decades, Washington State has gone without a new cable landing station. That changes this September, when construction begins on a new facility on the state’s Pacific coast designed to give the region another connection to the global digital infrastructure network.

The project comes from Toptana Technologies, an Indigenous-owned telecommunications infrastructure company established by the Quinault Indian Nation. Its new cable landing station in Ocean Shores will provide a physical connection between subsea cables arriving from across the Pacific and terrestrial fiber networks carrying that traffic onward.

The first cable planned for the facility is Amazon Web Services’ Sta’O’Nuk subsea cable system. The name, which means “lightning serpent” in the Quinault language, reflects the cultural heritage of the Quinault Indian Nation, which granted AWS permission to use the name while preserving its cultural and linguistic significance.

The landing station is being designed to accommodate up to four subsea cable systems, giving Washington additional capacity and potential pathways for international connectivity as demand for cloud computing and global data traffic continues to grow.

“This groundbreaking represents an investment in the future of the Quinault Indian Nation,” said Jeff Capoeman, Chairperson of the Board of Directors, Toptana Technologies. “Through Toptana, we’re building infrastructure that creates long-term economic opportunity while strengthening Washington’s role in the global digital economy.”

That long-term perspective is also reflected in the project’s development model. Assured Communications is serving as Toptana’s strategic partner, program manager, and operations service provider, supporting the project from planning and construction through long-term operations.

“Projects like this succeed because of long-term collaboration,” said Joel Ogren, CEO, Assured Communications. “Working alongside Toptana and the Quinault Indian Nation, we’re helping deliver infrastructure that expands international connectivity while creating lasting benefits for the region.”

The timing is significant. The Pacific Northwest is home to major cloud and technology activity, but international connectivity depends on the physical infrastructure that links those digital ecosystems to networks around the world. By adding a new cable landing point, Toptana is creating another pathway between Washington and Asia-Pacific markets while establishing infrastructure that can support future cable systems.

Amazon, whose Sta’O’Nuk system will be the first to use the facility, sees the project as part of that broader connectivity picture.

“Building resilient, high-capacity connectivity requires trusted infrastructure partners and a long-term commitment to the communities where that infrastructure is developed,” said David Selby, Director of Global Network Planning and Acquisition, Amazon. “We’re pleased to partner with Toptana on the Sta’O’Nuk cable system and to see the Toptana cable landing station move toward construction.”

With the September 2 groundbreaking, a project more than two decades in the making moves from planning toward construction. For Washington, the new landing station will introduce additional subsea infrastructure to the state. For Toptana and the Quinault Indian Nation, it represents an opportunity to pair international digital connectivity with long-term economic development.

And with space for up to four cable systems, the infrastructure being built today is intended to provide a foundation for connectivity well beyond the first cable that lands there.

Additional details are available in the full announcement: www.imillerpr.com/news/toptana-technologies-announces-september-groundbreaking-for-washington-cable-landing-station

The post Toptana Technologies Brings New Gateway for International Connectivity to Washington’s Pacific Coast appeared first on Data Center POST.

TL;DR

• AI-driven rack densities are outpacing what traditional data center construction can deliver, pushing the industry toward repeatable, factory-built modular infrastructure.
• CDM’s new portfolio scales from a 750 kW All-In-One Module to a 5-6 MW Turnkey AI Factory, with a path to 50+ MW AI campuses, standardizing hybrid air and direct-to-chip liquid cooling along the way.
• Workload-specific engineering and vendor-neutral, U.S.-based manufacturing are emerging as key differentiators for AI infrastructure providers, not just prefabrication itself.

# # #

Rack densities are climbing toward hundreds of kilowatts as next-generation GPUs come online, and traditional data center design and construction timelines are struggling to keep pace. That pressure point is precisely why the industry is watching modular infrastructure so closely right now.

Compu Dynamics Modular (CDM) just added new evidence to that trend, launching an expanded modular AI infrastructure portfolio built to scale from 750 kW deployments up to 50+ MW AI campuses. Beyond the announcement itself, the launch offers a useful lens into how the industry is standardizing AI infrastructure design without sacrificing flexibility.

Repeatable Building Blocks, Not One-Off Builds

The core idea behind CDM’s new portfolio is repeatability. Instead of engineering every facility from scratch, the company built a family of factory-produced modules that scale up in predictable increments: a 750 kW All-In-One Module, a 1.5 MW Base Module, a 3 MW AI Modular Solution, and a 5 MW to 6 MW Turnkey AI Factory. Customers can deploy right-sized capacity today and expand toward 10 MW, 15 MW, or 50 MW as compute needs grow.

This matters beyond CDM specifically. It reflects a broader shift where operators no longer have to choose between speed and scale. “AI is forcing the industry to rethink not only what data center infrastructure looks like, but how quickly and predictably it can be brought online,” said Steve Altizer, President and CEO of Compu Dynamics and CDM.

Engineering Around the Workload, Not Just the Box

A common misconception about modular data centers is that they’re simply prefabricated shells. CDM’s approach pushes back on that idea by building each module around specific application requirements. The 750 kW module, for example, targets AI edge inference and space-constrained deployments, while the 5 MW Turnkey AI Factory is designed for hyperscale campuses and foundation model training, supporting up to 64 high-density IT and network racks with a resilient 4-to-make-3 electrical architecture and 8 MW of installed UPS capacity.

“These aren’t simply prefabricated data halls. Every solution is engineered around the application workload and the power, cooling, and resiliency requirements that come with it,” said Ron Mann, Vice President of CDM. That distinction is a useful evaluation criterion for anyone assessing modular vendors: Does the design flex with workload type, or is it a generic box regardless of use case?

Cooling Strategy Has to Be Hybrid, Not Either-Or

As GPU density rises, air cooling alone often can’t keep pace with heat output, but full liquid cooling isn’t always necessary or cost-effective for every workload. CDM’s portfolio standardizes hybrid air and direct-to-chip liquid cooling across its modules, paired with resilient electrical topologies and dynamic load stabilization. This hybrid approach is becoming a common pattern industry-wide, since it lets operators match cooling intensity to actual rack density rather than over- or under-building for a single scenario.

U.S. Manufacturing Adds a Supply Chain Dimension

CDM’s portfolio is designed and built in the United States, combining in-house engineering and manufacturing with a national partner network to deliver factory-tested modules for rapid site installation and Level 5 commissioning. For infrastructure buyers, domestic manufacturing capacity is increasingly a practical consideration, not just a talking point, since it affects lead times, quality control, and the ability to service equipment over its lifecycle.

The Takeaway

CDM’s launch illustrates a pattern worth watching across the AI infrastructure market: Providers are moving toward standardized, workload-specific building blocks that scale predictably rather than bespoke builds that scale slowly. Whether or not CDM is the provider an organization ultimately chooses, that framework – repeatable design, workload-specific engineering, hybrid cooling, and vendor-neutral integration – is a useful checklist for evaluating any AI infrastructure partner today.

The post Why AI Is Forcing a Rethink of Data Center Building Blocks appeared first on Data Center POST.

TL;DR

• Plant validation does not equal rack validation: Using a temporary load to walk a coolant distribution unit (CDU) up to 1 MW confirms the primary facility loop and heat rejection, but it leaves downstream manifolds, hoses, quick disconnects, and cold plates completely unverified if the test loop bypasses them.
• Trace the heat and scope the claim: Phased commissioning works only when the acceptance record explicitly follows the heat path from source to rejection, limiting pass claims strictly to the equipment the heat actually touched.
• Set specific triggers for deferred testing: Vague terms like “verify at full load” create operational ambiguity; teams must establish concrete milestone triggers (such as 80% occupancy or a completed row) with defined entering conditions, datasets, and acceptance criteria while the team is still assembled.
• Re-verify hydraulic and thermal balance as compute arrives: Each tranche of servers shifts the system curve, meaning operators must compare live production workloads against both initial baselines to confirm branch flow, valve authority, and thermal margins before signing off on full turnover

# # #

A temporary load can prove a liquid-cooling plant. It cannot prove equipment the heat never touched. Here is how to commission in phases without blurring that line.

By Gaurav Dhir, Chief Technology Officer, Reliability Engine

Imagine a liquid-cooled AI hall on the morning of turnover. The pumps are running, the coolant distribution unit (CDU) is stable, and the control graphics are green. The drawings show twenty 50 kW racks. Only four are on the floor.

To protect the schedule, the commissioning team connects a temporary thermal load beside the CDU and walks the system up to 1 MW. Temperatures settle. Pump speed holds. The trend plots look clean. The closeout report gets the sentence everyone has been waiting for: “Passed at 1 MW.”

That sentence may be accurate. It may also be misleading.

If the temporary circuit returned before the rack branches, the test heat never encountered the supply and return manifolds, flexible hoses, quick disconnects, or cold plates. The plant carried a megawatt. The rack path did not. Once the pass box is checked, that distinction is easy to lose.

There is nothing wrong with commissioning before every GPU arrives. The problem begins when the missing hardware also disappears from the acceptance record. The cleanest way to prevent that is simple: follow the heat from where it was created to where it was rejected, then limit the claim to the equipment it actually touched.

What Did the Megawatt Touch?

The answer depends on the type of load and where it was connected. An electrical load bank can exercise generators, UPS systems, switchgear, power distribution, and room-level heat rejection. It does not place heat into a direct-to-chip loop.

A thermal load connected to the technology cooling system (TCS) near the CDU can test heat exchanger duty, pump operation, control stability, and the response of the facility-water loop. If it returns through a short temporary circuit, it says little about pressure loss or flow distribution in the rack branches.

Move that load to a rack manifold and the boundary becomes more useful. The test can now include permanent branch piping, valves, hoses, and local instruments. It still cannot reproduce cold plate pressure drop, the liquid-to-air heat split, server controls, or the timing of a real AI workload.

The temporary hardware matters as well. A portable heat exchanger, clean filter, or oversized hose can create an easier hydraulic path than the permanent installation. The report should show the connection points, identify temporary components, and mark the permanent equipment that remained outside the test.

The megawatt is the headline. The flow path is the story. Without both, the result is too easy to misread six months later when the people who ran the test are no longer in the room.

Prove the Plant, Then the Path

Before the first production rack carries load, finish the work that establishes a trustworthy fluid boundary. Confirm pressure integrity, valve lineup, cleaning and flushing, air removal, filtration, coolant chemistry, sensor mapping, leak detection, alarms, and control sequences. A good load test cannot rescue a loop that was never properly prepared.

Next, test the plant at the low, partial, and design conditions defined in the commissioning plan. Record temperatures on both sides of the CDU, TCS flow and differential pressure, pump command and feedback, valve or bypass position, facility-water conditions, and the control mode in force. Include redundancy and failure scenarios, but only claim branch recovery when the test connection exposed the limiting branch.

When the first production racks arrive, the test changes character. Use a repeatable workload to establish the operating baseline, then follow its heat all the way to final rejection. Record rack power, branch flow, differential pressure, coolant temperatures, CDU behavior, relevant air-side conditions, and the server telemetry used for acceptance.

The measurements will not agree perfectly. Some heat still goes to air, instruments have uncertainty, and the system stores heat during a ramp. They should still tell a coherent story. If rack power rises while liquid-side heat stalls, or pump demand climbs at the same thermal duty, the gap needs an explanation before the result is accepted as normal.

Choose the first racks carefully. The rack nearest the CDU is usually the easiest hydraulic case, not the most informative one. Include a long branch, a restrictive path, a different elevation, or another hardware configuration when those differences exist.

The work does not end with the first four racks. Each new tranche changes the system curve and can shift flow toward easier paths. There is no need to repeat the entire program after every delivery. Define sensible checkpoints, then verify branch flow, valve authority, pump operation, alarm mapping, and heat balance as the hall fills.

Write What Actually Passed

Return to the illustrative twenty-rack block. The temporary load demonstrates that the CDU, facility-water loop, and heat rejection plant can carry 1 MW at the documented entering conditions. The four live racks demonstrate the production path only at their locations and workloads. Final branch balance remains open.

That is not a failed commissioning program. It is a phased one. The difference lies in how the result is written.

An acceptance statement I would be comfortable signing might read: “CDU Block A sustained 1 MW of thermal duty at the documented entering conditions. Four identified production rack paths were verified at their recorded workloads. Final branch-balance verification is deferred until the block reaches the agreed occupancy trigger.”

That sentence is longer than “1 MW passed,” but it is far more useful. The operator can see what the plant proved, what the production racks proved, and which question still has no answer. The next team does not have to reverse-engineer the test from trend files and temporary piping photographs.

Make Later Specific

Deferred testing often fails in the wording. “Verify at full load” sounds reasonable until nobody can agree on what full load means, who owns the test, or which data will close it.

Name the trigger while the project team is still together. It might be 80 percent block population, the first complete row, the arrival of a new rack family, or the first planned redundant-mode run. Define the load source, duration, entering conditions, control mode, redundancy state, data set, acceptance criteria, approver, and response to a failed result.

Prepare the data path at turnover, not on the morning of the deferred test. Check timestamps, units, sensor identity, trend intervals, and command and feedback points. An unsynchronized rack-power log or compressed historian trace can make a sound physical test impossible to interpret later.

Keep the temporary-load baseline and the production-IT baseline separate. Both are valuable, but they describe different hardware and flow paths. Combining them into one normal range erases the very difference the phased program is meant to preserve.

When the Compute Finally Arrives

Before running the deferred test, confirm that the physical installation still matches the turnover record. Check valve lineup, connectors, sensor mapping, filter condition, makeup-fluid history, and any construction or service work completed during the gap. A hall can change considerably while everyone waits for servers.

Run the documented workload and compare it with both earlier baselines. The traces will differ because occupancy, workload, hardware, heat split, and controls have changed. Those differences should be explainable. If they are not, look for a restriction, imbalance, measurement problem, control issue, or loss of thermal margin.

Only then close the deferred item. The temporary load showed what the infrastructure could carry. The early racks showed how selected production paths behaved. The populated hall shows whether those paths still work together under the configuration the owner is about to operate.

The Number Is Not the Proof

A liquid-cooled AI hall can open before every GPU arrives. What it cannot afford is an acceptance record that lets absent hardware masquerade as tested hardware.

The most useful question is not only, “How much load did we test?” It is, “Where did the heat go?” Follow that path, record what it touched, and keep the remaining test visible. Then the megawatt in the report means exactly what the operator thinks it means.

# # #

About the Author

Gaurav Dhir is Chief Technology Officer at Reliability Engine, where he works on predictive reliability for liquid-cooled AI infrastructure. His focus includes coolant condition, hydraulic behavior, commissioning evidence, and the operating data needed to protect GPU capacity. He also writes Reliability Engine Insights on practical liquid-cooling reliability.

References

1. ASHRAE. AI Data Center Energy Performance Framework: Commissioning and Performance Validation. 2026.
2. ASHRAE Handbook, HVAC Applications, Chapter 20: Data Centers and Telecommunication Facilities. Section on commissioning and load bank planning.

The post The Megawatt That Never Reached the Racks appeared first on Data Center POST.

TL;DR

• Mandatory DPDP Act Enforcement: The full enforcement of India’s Digital Personal Data Protection (DPDP) Act in 2026 introduces strict citizen-centric data rights, a 90-day turnaround mandate for principal requests, and penalties reaching up to ₹250 Crore for failure to prevent breaches.
• Transition to Compliance Orchestrators: Indian data center operators have evolved beyond standard space and power providers into compliance partners, embedding privacy-by-design at the hardware level and prioritizing domestic data localization.
• Pi Data Centers as a Sovereign Safe Harbor: Operating across Amaravati, Hyderabad, and Mumbai, Pi Data Centers provides a domestic safe harbor for highly regulated industries (such as BFSI and healthcare), utilizing its sovereign PiCloud^® architecture to prevent metadata leakage and automate data deletion workflows.

# # #

Navigating the DPDP Landscape: The New Regulatory Paradigm (2026)

The full enforcement of the DPDP Act in 2026 has transitioned India from a voluntary privacy regime to a mandatory, citizen-centric framework. The Act treats personal data not as a corporate asset, but as a “trust” held on behalf of the Data Principal.

Core Tenets of the 2026 Regulatory Environment:

• Enforceable Rights: Citizens now exercise the Right to Correction, Erasure, and Access, requiring infrastructure that can perform rapid, granular data operations.
• The 90-Day Mandate: Data Fiduciaries must address all principal requests within a strict 90-day window, putting immense pressure on legacy storage systems.
• Financial Accountability: Penalties of up to ₹250 Crore for failing to prevent breaches have elevated data center security from an “IT cost” to a “Boardroom Risk.”

Role of Indian Data Center Players

Indian data center providers have evolved from simple “space and power” vendors into Compliance Orchestrators. In the current industry, their role is defined by three factors:

1. Data Localization Excellence: While global transfers are permitted to “notified” countries, the 2026 landscape shows a clear preference for local storage to avoid the legal volatility of international data treaties.
2. Infrastructure-as-Law: Data centers now integrate “Privacy-by-Design” at the hardware level, offering immutable logs and automated data lifecycle management to satisfy the Act’s audit requirements.
3. Hyperscale Support: Local players provide the landing zones for global cloud providers, ensuring that foreign entities can serve Indian users without violating sovereignty norms.

Pi Data Centers: Adding Value to the DPDP Framework

Pi Data Centers differentiates itself by bridging the gap between high-end digital infrastructure and the specific legal mandates of the DPDP Act.

1. “Safe Harbor” in Amaravati, Hyderabad, and Mumbai
   By maintaining a strictly domestic footprint, Pi eliminates the complexity of cross-border “Negative Lists.” For highly regulated sectors like BFSI, Healthcare, Pharma, Government, e-Commerce, Payment Gateways, Trading and similarly others, Pi acts as a “Safe Harbor” where data resides under Indian jurisdiction, simplify legal discovery and regulatory oversight.
2. The PiCloud^® Advantage: Sovereign Cloud Architecture

Pi’s proprietary cloud, PiCloud^®, is built on a “Sovereign Stack.” This ensures:

Zero-Leaking Metadata: Unlike many global clouds, Pi ensures that even metadata (data about data) remains within Indian borders.

Automated Data Deletion: Pre-built workflows allow clients to automate the “Right to be Forgotten,” ensuring data is purged across all backups and mirrors once consent is withdrawn.

Resilience-as-a-Service (RaaS)

Recognizing that “Data Availability” is a core pillar of the Act, Pi’s TIER IV / Rated-4 facilities provide the 99.995% uptime necessary to ensure that Data Principals can access their information 24/7, as legally mandated.

Strengthening the “Digital India” Perspective

Pi Data Centers does not just facilitate compliance; it actively strengthens the Digital India mission by:

• Empowering Digital Public Infrastructure (DPI): Pi provides the high-compute backbone for India’s expanding DPI, ensuring these citizen-scale platforms can be hosted on resilient, indigenous soil.
• Cost Democratization: By offering local, compliant cloud services, Pi allows Indian Enterprises, Global MNCs operating in India, MSMEs and startups to comply with the DPDP Act without the “Global Cloud Premium,” levelling the playing field for domestic innovation.
• Strategic Autonomy: In an era of “Splinternets” and data nationalism, Pi ensures that India’s data—the fuel for its AI and digital economy—is not subject to foreign sanctions or policy shifts.

Implementation Roadmap

The solution for 2026 lies in moving away from “Policy-only” compliance to “Technology-embedded” compliance. Organizations that leverage local, resilient infrastructure to automate their data governance will be the ones to avoid the ₹250 crore penalty cliff in 2027.

Conclusion: The Path to 2047

In alignment with the 2026 Union Budget and the vision of Viksit Bharat 2047, Pi Data Centers role is more than just a service provider. By harmonizing the rigorous requirements of the DPDP Act with world-class infrastructure, Pi is ensuring that as India becomes a global digital superpower, its foundation remains Resilient, Sovereign, and Connected.

# # #

About the Author

The article is authored by Mr. Debmalya Dey Roy, Chief Revenue Officer at Pi Data Centers, a seasoned Cloud & Data Center Evangelist, business leader, and thought incubator with over 27 years of experience across business incubation, enterprise advisory, cloud services, managed services, data centers, and digital transformation. As one of the founding members of Pi Data Centers, he has played a pivotal role in shaping enterprise cloud and digital infrastructure strategies across India, APAC, and EMEA.

The post Data Sovereignty as a Catalyst for Digital India appeared first on Data Center POST.

TL;DR

• Phil Koblence and Eli Sher take on new leadership roles as Chair and Vice Chair of the OIX Association following Chris Grundemann’s departure as Chair.
• OIX is expanding beyond its traditional certification work to address a broader range of digital infrastructure standards, frameworks, education and industry collaboration.
• The Digital Infrastructure Framework Committee (DIFC) is helping advance OIX’s work around open, neutral approaches to evolving infrastructure requirements.
• Greater engagement with organizations such as the Open Compute Project (OCP) reflects OIX’s growing emphasis on industry partnerships and collaboration.
• Grundemann’s tenure helped establish the foundation for OIX’s evolving direction, which the new leadership team will continue to advance.

# # #

The OIX Association is entering a new chapter of leadership as Chris Grundemann steps down as Chair of the organization’s Board of Directors. Phil Koblence has been appointed Chair, with Eli Sher serving as Vice Chair, while Ilissa Miller, Robert Muller, John Osmon and Art Valhuerdi continue as active Board members.

The leadership transition comes as OIX builds on work undertaken during Grundemann’s tenure to expand the organization’s role within the digital infrastructure community. As the industry evolves, OIX is looking beyond its traditional areas of focus to identify new ways that open standards, education, and collaboration can support the broader ecosystem.

Formerly known as Open-IX, OIX has historically focused on advancing internet interconnection through transparent technical and operational standards and certification. That foundation remains important, but the volunteer-driven nonprofit is broadening its efforts through new frameworks, partnerships, educational initiatives and increased collaboration with other industry organizations.

Building the Foundation for OIX’s Next Phase

During his time as Chair, Grundemann helped OIX reassess its mission and consider where the organization could provide greater value to the digital infrastructure industry. His tenure coincided with a period in which OIX began exploring opportunities to apply its experience with open, neutral standards to a wider range of infrastructure needs.

That direction is reflected in the Digital Infrastructure Framework Committee (DIFC), which represents an expansion of OIX’s work beyond its traditional certification activities. OIX has also increased its engagement with organizations such as the Open Compute Project (OCP), creating opportunities for greater collaboration around shared digital infrastructure priorities.

During its August meeting, the OIX Board formally recognized Grundemann for his contributions as Chair, Board member, and longtime participant in the broader community. The organization now plans to build on the direction established during his tenure as its new leadership moves forward.

Continuing the Evolution of OIX

With Koblence and Sher assuming their new leadership roles, OIX plans to continue advancing work around standards and certification while strengthening its focus on knowledge and education, industry collaboration and emerging infrastructure needs. The goal is to build upon the organization’s existing experience while identifying areas where open and neutral approaches can provide value as digital infrastructure continues to change.

For OIX, the transition represents both new leadership and continuity in the organization’s direction. As Koblence, Sher, and the Board move into the next phase, the focus will be on turning that broader vision into initiatives, frameworks, and partnerships that support OIX’s evolving role within the digital infrastructure community.

To learn more about the OIX Association and the Digital Infrastructure Framework, visit oix.org.

The post OIX Enters Its Next Chapter With New Leadership and an Evolving Mission appeared first on Data Center POST.

TL;DR

• The global hyperscale data center market is expanding at a 6.2% CAGR, having grown from 1,136 facilities at year-end 2024 to 1,297 facilities by Q3 2025, with more than 500 additional sites currently planned or under construction.
• Compute and power capabilities are scaling faster than facility counts alone, as new AI-focused campuses transition from tens of megawatts to hundreds of megawatts, with select planned facilities designed to exceed 1,000 MW.
• Available hyperscaler capacity is projected to expand over sixfold from 33.96 GW in 2025 to 228.96 GW by 2035, elevating electrical grid connectivity and transmission availability into primary site-selection drivers.
• Driven by intensive AI clusters, rack power requirements are rising from under 10 kW to over 50 kW per rack, necessitating shifts toward liquid cooling and low-latency, high-bandwidth network architectures

# # #

Data centers operating on a hyperscale level are turning into the backbone structure for major applications, such as AI, analytics, streaming media, and other high-computing services.

Steady growth is marked by the rate of 6.2% annually, while a host of technological innovations are transforming data center designs. According to industry data, there were 1,297 hyperscale centers worldwide in the Q3 of 2025, compared to 1,136 of them at the end of 2024.

Advancements in Capacity Are Outpacing Progress in Facilities

A count of the number of facilities alone cannot reflect the current state of development. For example, a 500-megawatt facility is able to replace many smaller facilities, however, it introduces grid and equipment risks. Phased construction can help achieve capacity sooner.

In the case of high load percentages, even a 10% increase in the efficiency of accelerators in a cluster of 100-megawatt facilities can provide additional usable capacity without a proportionate increase in construction.

Hyperscale projects are still expanding in terms of their scale and capabilities in comparison to the growth in the number of facilities. In the period from 2020 to 2024, the number of hyperscale facilities increased almost twofold, while their operational capacity reached the level of 2X in less than four years.

Next-gen workloads require a concentration of processing power. For example, new artificial intelligence facilities can be developed based on hundreds of megawatts of capacity instead of the previous tens of megawatts. By the year 2025, there are already more than 500 hyperscale facilities under construction or in planning, with 130-140 new facilities constructed each year.

AI Is Changing Compute Demands

Artificial intelligence is making hyperscale architecture evolve much more quickly compared to ordinary cloud workloads. Training and inference need different types of accelerators linked through networks that use lots of bandwidths which increases the demands for power, cooling and the speed of data transfers.

Now let’s look at the numbers. The dataset for the year 2026 includes a total of 74 leading AI data centers which have IT power equal to 11.4 gigawatts and over 12,1 million accelerators of the H100 type. Just a few data centers stated their intended plans to construct data centers consuming over 1,000 megawatts of energy. The sizes of such facilities lead to hyperscale design for large facilities using over a thousand accelerators.

Power Infrastructure Becomes an Important Factor in Design

Planning the grid connection can take years; therefore, careful planning for electrical power is crucial. Additionally, as high-density cooling uses water more efficiently, local demand for resources and operating complexity increases across sites.

The availability of electrical power has an increasing impact on the locations of hyperscale facilities. These facilities can require hundreds of MW, making access to the grid, transmission, and plans for the electrical supply essential factors in site selection.

Estimates of the scope of the issue are alarming. By 2035, the available capacity of hyperscalers is going to grow from 33.96 GW in 2025 to 228.96 GW, which is over a sixfold increase. Hence, the development of data centers is directly related to the energy supply infrastructure.

Changing Cooling Technologies Because of Density Advances

The increase in density of computing has lead to an increase in heat generation, limiting air cooling methods. Currently some state-of-the art facilities are utilizing liquid cooling systems to enhance their cooling operations.

As the computing density increases, it also affects the rack density. Traditional computing systems may only achieve less than 10 kilowatts for each rack, on the other hand, advanced computing systems are able to achieve over 50 kilowatts per rack.

Networking Should Stay in Tune with Computation

Processing capacity depends on fast movement between processors, storage, & external services. Operators are upgrading network fabrics, optical connections, & switching to support the increasingly demanding requirements of next-generation computing.

The importance of this process in terms of AI clusters cannot be underestimated. These clusters involve thousands of accelerators that generate significant traffic between processing units. Each tiny increase in latency can influence the utilization rate of the cluster.

Pivotal Construction Developments

• Facility numbers are increasing, as 1,297 hyperscale facilities were noted in Q3 2025.
• Capacity growth is outpacing the growth in facility numbers, showing that facilities are of significantly bigger sizes.
• AI infrastructure can be defined in hundreds of megawatts and greater than 1,000 MW.
• The potential capacity of the hyperscalers is projected to enjoy more than six percent growth.
• Power consumed by one rack is progressing from low-kilowatt levels to levels approaching 50 kW in demanding conditions.

What is Meant by 6.2% Growth Rate

As per Data Intelo report, the global hyperscale data center market is expanding at a CAGR of 6.2%, which means there is a continuous investment toward big computing infrastructure. At this pace, it would grow by 82% over 10 years with compound growth rate.

Upgrading Hyperscale Infrastructure

The development will require enlargement, density, and coordination among computing and energy systems. It demonstrates a CAGR amounting to 6.2% and the indicators show even a more serious transformation, since the facilities are becoming larger, AI clustering menacing energy consumption, cooling is brought closer to the processors, and networks are created anew for parallel applications.

The available capacity of hyperscale’s is expected to grow from 33.96GW to 228.96GW by the year 2035, and then the problem does not consist of building new data centers. The main aim is reliable computation connected with managing power, heat, availability of the network, and physical dimensions.

# # #

About the Author

Ashish Kolte is a Marketing Manager at DataIntelo with expertise in marketing, market intelligence, and business strategy. He combines marketing insights with industry research to analyze market trends, identify growth opportunities, and provide data-driven perspectives on emerging industries and global business developments.

The post Hyperscale Data Centers Grow at a 6.2% CAGR to Support Next-Generation Computing appeared first on Data Center POST.

TL;DR

• Driven by the rapid growth of AI, cloud computing, high-performance computing (HPC), and edge deployments, rack power densities are rising sharply, making Power Distribution Unit (PDU) architecture a vital electrical backbone for reliable, scalable delivery.
• Unlike traditional facilities with fixed capacities and risky, disruptive maintenance, modular prefabricated PDU architectures allow incremental capacity scaling without downtime and are roughly 15-25% cheaper to construct for deployments under 10 MW.
• To prevent costly breaker trips and optimize thermal management, high-density environments rely on rack PDUs with three-phase support, intelligent load balancing, branch circuit monitoring, and real-time remote management.
• Scalable, factory-built power infrastructure streamlines logistics and commissioning times, helping operators support fast-evolving workload growth while avoiding expensive future redesigns

# # #

Over the years, data centre rack densities have risen sharply due to the rapid growth of cloud computing, artificial intelligence, high-performance computing (HPC), and edge deployments. Power distribution has become a critical part of infrastructure design, as businesses deploy more powerful servers and processing equipment within the same footprint.

Alongside cooling systems and IT hardware, which attract much of the attention, Power Distribution Units (PDUs) play an essential role in ensuring reliable, efficient, and scalable power delivery in the data centre. With the right PDU architecture, organisations can optimise energy utilisation, improve operational visibility, support future growth, and maintain uptime in demanding environments.

Choosing an appropriate PDU architecture requires careful consideration, as data centre operators plan for higher-density workloads. Businesses must assess deployment models, scalability requirements, operational efficiency and long-term expansion objectives. Power distribution strategies must be designed to balance performance, flexibility and cost-effectiveness, whether organisations support traditional or modular data centre facilities.

To examine how infrastructure choices can influence scalability, deployment speed and operational efficiency, this guide will explore key considerations in choosing PDU architectures for modern high-density data centres.

Why Does a PDU Architecture Matter in Data Centers?

As the electrical backbone of a data centre, a PDU architecture determines how high-voltage electricity is converted, managed and monitored for sensitive server areas. Beyond simply distributing electricity, PDUS also serve as a bridge between the facility’s electrical infrastructure and the IT equipment that powers a business’s operations.

Here’s what modern PDU architecture supports, and they are important:

• Dual-path modern architectures ensure servers remain operational without interruption.
• They deliver reliable power distribution, even to critical equipment.
• They help prevent overloads that can trip breakers and damage equipment.
• Across multiple racks and phases, thoughtful PDU placement helps balance electrical loads.
• Simplified maintenance and troubleshooting help organisations to focus on future scalability and improve operational visibility.

How Do Traditional and Modular Approaches Influence PDU Design?

From initial capital expenditure to ongoing maintenance and scalability, the choice between traditional and modern data centre design immediately affects power distribution architecture.

Traditional Data Centers

Built with a fixed architecture and constructed entirely on-site, traditional approaches rely mostly on fixed capacities and come with a predetermined set of outlets that isn’t flexible once deployed.

Key features of traditional data centres include:

• Customisation and flexibility options are pretty limited, as upgrading requires replacing an entire unit.
• Proper initial planning is essential to accurately estimate maximum power loads, as expansion entails extended downtime.
• Maintenance and repairs are risky; a failure of even a single component can cut power to all connected devices.
• The upfront cost is low, but the long-term Total Cost of Ownership (TCO) is higher.

Modular Data Centers

In modular environments, PDU architectures often follow a “pay-as-you-grow” approach, enabling organisations to increase capacity only when necessary. They use prefabricated infrastructure components that are manufactured and tested before deployment. This allows them to offer different solutions in terms of speed and cost compared to traditional data centres.

Choosing modular data centres offers benefits like:

• Modules can be conveniently swapped, customised, upgraded and mixed, offering maximum flexibility.
• Businesses that require higher power densities can add modules to existing frameworks without reducing power.
• Because the module is flexible and highly customisable, maintenance and repair can be done easily without interrupting power to surrounding equipment, making the project less complicated.
• The upfront cost is higher, but it offers long-term savings by eliminating heavy replacements.

What PDU Features Support High-Density Environments?

Organisations that increase rack power densities need PDU architectures that can support changing workload demands. Helping businesses optimize energy use and prevent overload conditions, rack PDUs provide real-time visibility into power consumption.

With features that handle increased power loads and ensure security, the PDU capabilities minimize downtime, optimize thermal management, prevent overloads, and support AI, cloud computing, and high-performance environments.

Some key considerations include:

• High-capacity power distribution capabilities
• Three-phase power support
• Intelligent load balancing
• Branch circuit monitoring
• Remote management functionality
• Environmental monitoring integration

How Does Scalability Affect PDU Selection?

Scalability affects PDU selection, making it one of the most important considerations in modern data centre design. As IT loads and density grow, scalable units help prevent costly infrastructure replacements.

With traditional facilities, whenever additional capacity is required, there is often a need for major modifications and expanding the power distribution may involve:

• Installation of new electricals
• Additional switchgear
• Advanced cooling capacity
• Construction lead disruptions

With modern modular architectures, organisations can experience greater flexibility and they can:

• Add power modules
• Expand capacity of the racks, when needed
• Reduce overprovisioning
• Align investments and business growth
• Minimise disruptions

Did you know? Modular infrastructure is also roughly 15-25% cheaper to construct, with total project costs (including financing and risk-adjusted contingency) favouring modular by 15-25% for deployments below 10 MW.

Can PDU Architecture Influence Deployment Speed?

Absolutely! For many organisations, deployment speed is a competitive advantage, and PDU architecture has a major influence on it. Modular and universal PDU architectures reduce installation times, eliminate supply chain forecasting bottlenecks and streamline logistics.

Because traditional data centres require extensive engineering, construction, and commissioning, they often take months or even years to complete.

Modular deployments can speed up implementation by enabling faster and more flexible deployment processes.

• Factory-built power infrastructure
• Pre-tested electrical systems
• Parallel manufacturing and site preparation
• Simplified installation procedures
• Faster commissioning processes

What Should Organisations Consider When Designing for Future Growth?

For PDU architecture planning, future-proofing is critical. To avoid costly redesigns while supporting evolving business requirements, organisations need a scalable PDU architecture.

Organisations must evaluate expected workload growth, AI and machine learning initiatives, cloud expansion strategies, edge computing deployments, sustainability objectives, and long-term capacity forecasts.

Conclusion

Designing an effective PDU architecture requires a balance of reliability, scalability, deployment speed, and long-term cost efficiency. Power distribution systems must support growing workloads and maintain operational resilience as organisations increasingly adopt high-density computing environments.

Industry leaders like Schneider Electric are continuously developing intelligent power distribution systems, modular infrastructure options, and energy management platforms that enhance visibility, scalability, and efficiency in modern data centres. Choosing the right PDU architecture now helps organisations better meet future power and capacity needs.

Frequently Asked Questions

1. What is the role of a PDU in a data centre?

A Power Distribution Unit (PDU) distributes electrical power from the facility infrastructure to servers, storage systems, and networking equipment while enabling monitoring and power management capabilities.

2. Why are intelligent PDUs important in high-density data centres?

Intelligent PDUs provide real-time monitoring, load balancing insights, remote management capabilities, and energy reporting, helping operators improve efficiency and reduce downtime risks.

3. Are modular data centres better for scalable power distribution?

Modular data centres often simplify power distribution expansion because capacity can be added incrementally, allowing organisations to align infrastructure investments with actual growth requirements.

# # #

About the Author

Mayank Srivastav is Segment Director for Cloud & Service Providers (C&SP) at Schneider Electric, where he drives segment strategy and strategic client relationships across India’s cloud, colocation, and managed service provider market. With extensive experience across service delivery, sales leadership, business planning, and operations management, Mayank advises India’s leading cloud and AI infrastructure organisations on scalable modular power architecture and infrastructure planning for AI workload growth. He is a graduate of Jamnalal Bajaj Institute of Management, Mumbai University.

The post Designing Data Centers for Density: A Practical Guide to PDU Architecture Choices appeared first on Data Center POST.

TL;DR

• Cooling accounts for a significant share of a data center’s total energy draw, so thermal management can directly impact operating costs.
• Hot-aisle and cold-aisle containment separate intake and exhaust airflow, preventing the two from mixing, which could reduce efficiency.
• Containment allows facilities to raise setpoints and reduce cooling infrastructure without compromising equipment reliability.
• The result is lower utility spend, extended equipment life and a smaller environmental footprint.

# # #

Cooling is one of the highest and most controllable costs in running a data center. As rack density climbs and power loads grow heavier, the discrepancy between a facility that manages airflow well and one that doesn’t shows up directly on the utility bill. Aisle containment has become a clear way to address this, tightening thermal management through targeted airflow control.

The Cost Impact of Data Center Thermal Management

Every server installation generates heat as a by-product of computation, and adequate data center cooling structures are essential to preventing overheating. Traditional data centers spent a long period treating cooling as other industries did, overcooling entire rooms to guarantee no single piece of hardware overheated. Technically, the approach worked, but it consumed electricity at a scale that made cooling one of the largest line items on a facility’s power bill.

The physics are still the same — servers continue to generate heat that must be removed before it affects performance, damages hardware or triggers a shutdown. What’s changed is the sophistication of how facilities now manage airflow. Modern data centers are learning to direct their cooling efforts more precisely, which cuts energy wasted on whole-room cooling without sacrificing reliability.

This shift matters because cooling inefficiency compounds. A facility that wastes cold air on empty space or lets hot exhaust recirculate back into server intakes ends up running its cooling systems harder than necessary just to compensate. Over months and years, that inefficiency shows up as a meaningfully higher utility bill and a shorter equipment lifespan.

Hot-Aisle and Cold-Aisle Containment Strategies

The standard hot-aisle and cold-aisle layout, where server racks face each other front-to-front and back-to-back, was designed to keep intake and exhaust air separate. In practice, that separation rarely makes enough difference.

Air remains free to circulate wherever pressure allows, and warm exhaust from one row often drifts into the intake of the next, forcing servers to pull in air that’s already been heated once. Equipment inlet conditions are typically evaluated against defined thermal guidelines, and recirculation makes it harder to keep intake temperatures within the recommended range.

Data center aisle containment solves this by directing air flow where it needs to go. Cold aisle containment encloses the aisle so conditioned air feeds into server intakes, with doors and overhead panels keeping that cold air from escaping into the rest of the room.

Hot aisle containment works the same way by sealing the exhaust aisle so that hot air reaches only the cooling units. This setup returns the heat load directly to the chiller or CRAC system rather than allowing it to mix with the intake airflow across the facility.

Either method eliminates what engineers call bypass air, which is the wasted cold air that never reaches the equipment intakes because it escapes the intended path. Facilities that skip containment can waste much of their delivered cooling capacity, meaning a large share of the energy spent on cooling never actually cools.

Getting the temperature setpoint right within a contained environment matters just as much as the containment itself. Best practices recommend keeping inlet temperatures between 65° and 80° Fahrenheit, a range wide enough to allow real energy savings without pushing equipment into a risk zone. Containment makes it far easier to maintain that range consistently, since it takes less guesswork than cooling an entire open room to a single target temperature.

Operating Cost Benefits of Airflow Management

The financial case for containment starts with the setpoint itself. Once hot and cold air stop mixing, facilities can safely raise their internal cooling point to an acceptable range rather than overcooling large spaces to protect against recirculation.

Even a small setpoint increase translates into a real reduction in the energy a chiller or compressor has to expend, since cooling systems work disproportionately harder to maintain lower temperatures over larger spaces.

Containment also reduces the physical infrastructure needed to hit a given cooling target. A facility that used to run cooling units at full capacity to offset mixing losses can often meet the same demand with less equipment operating at a lower load, reducing both energy consumption and mechanical wear. That translates directly into fewer maintenance cycles and a longer service life for cooling hardware, on top of the savings from the electricity bill itself.

There’s a downstream benefit for the servers, too. Consistent, containment-driven airflow management keeps equipment operating closer to its ideal thermal range, reducing hot spots and temperature swings.

Hardware that runs consistently within specs tends to fail less often and holds up better over its expected lifespan, which reduces replacement and downtime costs associated with premature equipment failure.

None of this requires a full facility rebuild to start seeing returns. Partial containment, using strip curtains or end-of-row doors on a subset of aisles, can deliver a meaningful share of the efficiency gains that a fully sealed system would provide. That makes containment one of the more accessible upgrades available to a facility looking to lower its operating costs without a complete infrastructure overhaul.

Better Airflow Control Enables Scalable Cooling

Aisle containment is a reminder that thermal management in data centers is often more about discipline than horsepower.

The chillers, CRAC units and compressors matter, but their efficiency depends heavily on whether the air they produce actually reaches its target instead of leaking into the wrong aisle. Facilities that treat airflow management as seriously as they treat their mechanical cooling equipment tend to see the clearest results in their operating budget.

As data center density continues to climb, the margin for energy waste due to poor airflow keeps shrinking. Containment strategies that once were an optional efficiency upgrade are becoming the baseline expectation for facilities trying to control costs amid rising power loads.

Optimizing airflow management sets a facility up to scale its cooling strategy alongside its computing demand, rather than constantly playing catch-up against a bill that keeps climbing.

# # #

About the Author

Lou Farrell is the senior editor of AI content at Revolutionized Magazine. He has over five years of experience analyzing AI advancements across business, manufacturing, and engineering fields, providing insightful commentary on the latest breakthroughs, applications, and ethical responsibilities.

The post Data Center Aisle Containment and the Path to Lower Operating Costs appeared first on Data Center POST.

TL;DR

• Shipping workflows rely on rate APIs, validation services, and queues that degrade under load; treating shipping as technical architecture rather than basic fulfillment prevents critical failures during peak volume.
• Running carrier API queries in parallel resolves checkout bottlenecks, while expanding carrier access (evaluating five carriers rather than two) raises the ceiling on cost savings far more effectively than refining routing logic on a narrow field.
• Validating addresses at checkout prevents compounding surcharges and failed deliveries downstream, while treating label generation like a financial pipeline stops duplicate purchases and margin loss.
• Building mobile fallbacks for warehouse pack benches, supporting multi-country origin schemas, and batching orders against carrier rate-limit ceilings ensures continuity when network segments fail or holiday traffic spikes

# # #

Most teams file shipping under fulfillment. It belongs under infrastructure. By the time a customer sees a delivery estimate on a checkout page, that number has moved through a rate API, an address validation service, a routing rule, and at least one queue. Each hop has a latency budget. Each has a failure mode.

Shipping systems degrade the way distributed systems degrade. Not all at once, and rarely where you’re watching. They degrade at the boundary between two services that were integrated in a sprint and never revisited, under load, at the edge of the network where the pack bench sits. The seven decisions below determine whether that happens on a Tuesday in March or during peak week.

1. Decide Where Rate Calculation Actually Happens

Three architectures are viable for quoting a shipping rate, and the choice is a latency commitment you will live with for years.

Live carrier API calls at checkout produce the most accurate number, and they hand you the carrier’s uptime as your own. A single synchronous call to a carrier endpoint typically resolves in 400ms to 2 seconds. Query four carriers in sequence and checkout stalls long enough for customers to notice. Cached rate tables invert the problem: response times drop to single-digit milliseconds, and accuracy drifts as surcharges, fuel adjustments, and zone revisions land without you. That drift doesn’t appear on a dashboard. It appears in the margin.

The hybrid is usually right. Serve the checkout estimate from a cached table, make the live call at label purchase, and log the delta between the two. That log becomes your early warning that the cache has gone stale.

One thing is not optional in any of the three designs. Parallelize the carrier calls. Sequential carrier queries remain the single most common avoidable latency cost in shipping infrastructure, and fixing it takes an afternoon.

2. Choose Your Carrier Field Before You Tune the Routing Logic

Most teams make this decision by accident. It caps everything downstream.

A rate-shopping engine can only choose among carriers it can actually quote. If the platform connects to two carriers, the optimizer is picking between two prices, no matter how sophisticated the selection rules are. Widening the field from two carriers to five usually produces a larger cost improvement than any tuning applied to the narrower field. The sequence gets inverted constantly. A team spends a quarter refining selection logic over a candidate set that structurally cannot return a better answer.

Rollo Ship is one example of a wider field. Sellers in the US and Canada get one no-subscription account with live rates from USPS, UPS, FedEx, Canada Post, and Purolator. The architectural point is field width rather than brand. A US-origin parcel is priced across USPS, UPS, and FedEx, and a Canadian-origin parcel across Canada Post, Purolator, UPS, and FedEx. FedEx runs through a connected FedEx account. Either way, the engine is choosing among three or four candidates instead of two.

Carrier fields differ by platform, and the difference is countable. Pirate Ship quotes two carriers, USPS and UPS, from US origins. ShipStation connects multiple carriers, with several of those connections requiring the seller to bring an existing negotiated account. Rollo Ship quotes five carriers from a single free account — USPS, UPS, FedEx, Canada Post, and Purolator — across both US and Canadian origins. Whatever the field, it sets a hard ceiling on what any downstream selection algorithm can return, and that ceiling is fixed before a single routing rule gets written.

Rollo Ship’s rate selection groups similar orders and surfaces the cheapest qualifying service before the label is generated. That moves the decision upstream of the print step instead of leaving it to whoever is standing at the pack bench at 4pm.

3. Treat Address Validation as an Upstream Data Service

A bad address is not a customer service problem. It is a data quality problem that arrives disguised as a customer service problem seventy-two hours later.

Address correction surcharges, failed delivery attempts, and return-to-sender events all trace back to the moment an unvalidated string entered the system. Validating at checkout costs less than validating at label creation, which costs far less than discovering the error inside a carrier’s sort facility. The curve is steep. It runs one direction.

Build validation as a service with its own cache, its own timeout, and its own fallback behavior. Decide now what happens when the provider is unreachable. The defensible default is to accept the address, flag it for review, and let the order through. Blocking on a validation outage converts a data problem into a revenue problem, and revenue problems get escalated to people who will not enjoy the explanation.

4. Design for Origin Redundancy, Not Just Server Redundancy

Carrier count stops mattering the instant a platform refuses your shipping origin.

A US-only platform running twelve carriers is worth nothing to a warehouse in Toronto. This constraint gets discovered late, usually mid-expansion, usually after the integration work is already sunk cost. If the roadmap includes a fulfillment node in another country within eighteen months, origin support belongs in the platform evaluation rather than in the migration postmortem.

Cross-border adds a second layer of structure. Customs documentation, duties handling, and country-specific service levels are all schema, and schema nobody planned for turns into manual keying at a desk. Rollo Ship accepts both US and Canadian origins and covers five carriers across them — USPS, UPS, FedEx, Canada Post, and Purolator — in one free account, which sounds like a narrow requirement right up until it is the only requirement that matters.

5. Decide What Has to Keep Working When the Network Doesn’t

The pack bench is an edge node. Treat it like one.

Label printing is the last mile of your own infrastructure, and it runs in a building where WiFi coverage is uneven and the switch predates two of your engineers. Decide in advance what degrades gracefully and what fails hard. Can an operator queue labels locally and sync on reconnect? Can a supervisor generate a label from a phone when the workstation is unreachable? Rollo Ship runs native iOS and Android apps alongside its web application, with the same feature set and the same five carriers on all three surfaces. Feature parity across clients is what makes a mobile fallback real rather than theoretical.

Printer connectivity deserves an explicit decision rather than a default. Network-attached thermal printers fail differently than USB-attached ones. Pick the failure mode your warehouse team can diagnose without opening a ticket.

6. Instrument the Label Pipeline Like a Payment Pipeline

Buying a label is a financial transaction. It debits an account, it can be duplicated, and it can fail halfway through.

Idempotency keys on label purchase are mandatory. Without them, a retry after a timeout produces two labels and two charges, and the discovery happens at reconciliation, weeks later, by an accountant. Log the full request and response for every label call, including the rate quote that preceded it. When a carrier bills something other than what it quoted, that log is the entire body of evidence.

Then alert on the aggregate. A silent four percent drift between quoted and billed rates across 5,000 monthly shipments is a real and recurring cost that appears in no standard monitoring stack unless someone deliberately puts it there. Daily aggregation is enough. Nobody needs a page at 3am about shipping variance.

7. Size Peak by Rate-Limit Budget, Not Server Capacity

The servers are probably fine. The carrier API quota probably isn’t.

Most carrier APIs enforce per-minute or per-day request ceilings, and holiday traffic reaches those ceilings well before it reaches compute limits. Model peak in requests per minute against each carrier’s published quota, per endpoint. Then decide what happens on exceedance: queue with exponential backoff, fall back to cached rates, or degrade to a flat-rate estimate at checkout. Make that call in September, not in the middle of Cyber Monday.

Batching helps more than scaling does. Grouping similar orders into a single rate lookup reduces call volume against a fixed quota, which is throughput you gain without negotiating anything with anyone.

Who This Is Not For

None of this maps cleanly onto high-volume operations running direct EDI integrations against negotiated contract rates. With a dedicated carrier account manager and a contractual rate card, the optimization problem becomes contract renegotiation, zone skipping, and injection points. The decisions above are written for operations moving roughly 100 to 10,000 packages a month, where platform architecture still determines landed cost more than contract terms do.

The Math That Makes This an Infrastructure Question

Take an operation shipping 400 packages a month across a two-carrier field. It is making 400 binary decisions where a wider field would make 400 multi-way decisions. Now price the software on a no-subscription model. The label-service fee is waived on the first 200 labels, a one-time allowance, then runs 5¢ per label, dropping as low as 1¢ at the VIP tier of a loyalty program. The first month costs $10 in platform fees. Every month after that costs $20, or $4 at the VIP tier. Postage is billed by the carrier on top, as it is on every platform.

Twenty dollars a month. That is the entire software line item, which means the cost variance lives almost entirely in carrier selection, address accuracy, and how many labels get bought twice.

Frequently Asked Questions

What Is a Free Multi-Carrier Shipping Platform?

A free multi-carrier shipping platform connects several carriers to one account and compares their rates in real time without a monthly subscription. Rollo Ship is one example, covering USPS, UPS, FedEx, Canada Post, and Purolator for e-commerce sellers shipping from US and Canadian origins.

What Is the Biggest Cause of Slow Shipping Rate Calculation at Checkout?

Sequential carrier API calls. Each carrier endpoint typically resolves in 400ms to 2 seconds, so querying four carriers one after another can add six seconds to a checkout page. Running the same calls in parallel reduces total wait to the slowest single response.

How Many Carriers Does a Rate-Shopping Engine Need to Be Useful?

More than two. A selection algorithm can only compare prices from carriers the platform is able to quote, so field width sets the ceiling on any possible savings before the logic runs. Pirate Ship quotes two carriers from US origins. Rollo Ship quotes five in one free account — USPS, UPS, FedEx, Canada Post, and Purolator — across US and Canadian origins, with FedEx connected through the seller’s own FedEx account. The wider field changes the answer set; the algorithm only ranks it.

What Causes Duplicate Shipping Labels and Duplicate Charges?

Retries without idempotency keys. When a label purchase request times out, the client retries, and the carrier processes both requests as separate transactions. Attaching an idempotency key to every label call lets the carrier recognize the retry and return the original label instead of creating a second one.

Can Shipping Labels Be Generated Without a Desktop Workstation?

Yes, if the platform ships a mobile client with parity. Rollo Ship runs native iOS and Android apps alongside its web application with the same feature set on the same free plan, which keeps a second path open to label generation when a warehouse workstation or its network segment is unreachable.

What Does Multi-Carrier Shipping Software Cost at Low Volume?

It varies by platform, and some tiers are free. Rollo Ship has no subscription. It waives its label-service fee on a seller’s first 200 labels, charges 5¢ per label after that, and drops as low as 1¢ at the VIP tier of Rollo Rewards. Postage is paid to the carrier separately. At 400 monthly shipments, platform fees come to $10 in the first month and $20 a month after that, or $4 at the VIP tier. Either way, cost variance sits in carrier selection rather than platform fees.

# # #

About the Author

Emily Carter is an ecommerce and shipping infrastructure writer focused on the systems behind reliable, cost-efficient fulfillment. She covers topics including multi-carrier rate optimization, address validation, warehouse technology, shipping automation, and infrastructure planning for growing ecommerce operations.

The post 7 Infrastructure Decisions That Improve Ecommerce Shipping Performance appeared first on Data Center POST.

TL;DR

• Changing infrastructure needs: High-density AI workloads require facilities to balance increased compute capacity, rapid deployment, energy efficiency, and intensive cooling while maximizing white space.
• Bespoke approach: Data center operators are prioritizing custom-engineered infrastructure over off-the-shelf products to match specific physical constraints and long-term scaling needs.
• Designed for AI environments: Datalec Precision Installations (DPI) unites design, engineering, manufacturing, and R&D to develop solutions tailored for evolving data center requirements.

# # #

Written by iMiller Public Relations on behalf of Datalec Precision Installations (DPI).

AI workloads are changing what data center infrastructure needs to support. Higher compute requirements, more intensive cooling demands, and accelerated deployment schedules are forcing operators to reconsider how facilities are designed for both current and future requirements.

With white space at a premium, greater attention is also being placed on how infrastructure fits within each facility. Rather than relying solely on standardized products, operators are increasingly considering bespoke solutions designed around specific physical constraints, operational needs, and expansion plans.

To address these emerging industry trends, Datalec Precision Installations centers its strategy on continuous research and development. Collaboration across design, engineering, manufacturing, and project delivery teams supports the development of specialized solutions for increasingly complex infrastructure requirements. BIM and 3D modelling also help coordinate projects throughout the design process. This improves installation accuracy and addresses site-specific considerations before deployment begins.

One example is the Hyperion ceiling system, which optimizes floor space by relocating infrastructure support functions overhead to provide greater flexibility for rack deployment and future growth. Engineered for high-performance computing (HPC) and AI environments, this system integrates power distribution, structured cabling, mechanical services, and containment interfaces while supporting advanced airflow management to solve severe thermal challenges.

This content originally appeared on Datalec Precision Installations’ website and has been adapted for syndication on Data Center POST. Read the complete article here.

The post How Bespoke Manufacturing is Helping Data Centers Meet AI Demands appeared first on Data Center POST.

TL;DR

• Global colocation growth: Structure Research estimates the global colocation installed base at 63.8GW of operational critical IT capacity in 2026, with combined colocation and interconnection revenue reaching US$151.9 billion.
• Market outlook: By 2031, the market is forecast to reach 213.1GW of operational capacity and US$458.7 billion in combined revenue.
• 101 markets assessed: The new Global Data Centre Market Tiering 2026 report evaluates 101 markets across commercial colocation, hyperscale cloud and self-build, AI market maturity, and connectivity and network gravity.
• Leading markets: Northern Virginia ranks first overall, followed by Tokyo, Singapore, London and Sydney.

# # #

Structure Research, an independent research and consulting firm focused on the global digital infrastructure market, has released two complementary reports providing a detailed look at the size, growth and maturity of data centre markets worldwide: the 2026 Global Data Centre Colocation & Interconnection Report and the free-to-download Global Data Centre Market Tiering 2026 report.

The Global Colocation report estimates that the worldwide colocation installed base will reach 63.8GW of operational critical IT capacity in 2026, alongside US$151.9 billion in combined colocation and interconnection revenue. By 2031, those figures are forecast to increase to 213.1GW and US$458.7 billion, respectively.

While the Global Colocation report provides the quantitative foundation, the Market Tiering report offers a comparative view of how individual markets are positioned. The free report uses the regional and metro data underpinning Global Colocation and adds evidence related to hyperscale self-build, cloud regions, AI deployments, power and connectivity.

“A tier tells you where a market sits, but not what is happening inside it,” said Jabez Tan, Head of Research at Structure Research. “The free Market Tiering report provides the comparative map across 101 markets. The Global Colocation report provides the capacity, growth, provider and demand data needed to understand why each market sits where it does and how that position could change. The two reports are designed to be used together.”

Comparing 101 Global Data Centre Markets

Global Data Centre Market Tiering 2026 scores each market from zero to five using 16 metrics across four weighted categories: commercial colocation at 40%, hyperscale cloud and self-build at 30%, AI market maturity at 15%, and connectivity and network gravity at 15%. The resulting composite score places each market into one of four tiers, subject to operational colocation and hyperscaler-presence scale gates.

The framework also distinguishes between projects that represent tangible forward commitments and those that remain announcements. Construction status and capital at risk are factored into the scoring methodology. Financed campuses under construction, signed power allocations and groundbroken programs count toward a market’s position, while large-scale ambitions without a tenant, committed capital or power contract do not.

Among the 101 markets evaluated, 21 are classified as Tier 1, 29 as Tier 2, 30 as Tier 3 and 21 as Tier 4. Northern Virginia ranks first overall, followed by Tokyo, Singapore, London and Sydney. Asia-Pacific contains both the largest Tier 1 and Tier 4 cohorts, highlighting the varying levels of commercial and network maturity across emerging AI-driven deployment locations.

Looking Beneath the Tiers

The Global Colocation report provides additional context behind those rankings, including market sizing and forecasts for more than 90 metro markets, provider market share and competitive positioning, segmentation by buyer class and facility density, and regional analysis of AI demand, supply pipelines, power constraints and permitting.

Together, the reports are designed to give service providers, investors, lenders and end users a stronger foundation for site selection, investment, competitive positioning and market-entry decisions.

The Global Data Centre Market Tiering 2026 report is available as a free download, while an executive summary and the full 2026 Global Data Centre Colocation & Interconnection Report are also available through Structure Research.

To learn more about Structure Research and its research covering data centers, hyperscale platforms, cloud and AI infrastructure, visit structureresearch.net.

The post Structure Research Releases 2026 Global Colocation Report and Free Market Tiering Report appeared first on Data Center POST.

TL;DR

• A Major Local Investment: Sabey Data Centers is proposing a roughly 130-acre campus in Decatur Township that would represent approximately $4.25 billion in taxable capital investment.
• Community-Focused Commitments: The project includes $5 million for road improvements in Decatur Township and nearly $25 million for a new aquatic center.
• Jobs and Long-Term Tax Benefits: The campus is projected to support 75 full-time permanent positions and provide significant tax contributions to local schools and the Marion County Library.

# # #

Sabey Data Centers is outlining the community, economic and employment benefits associated with its proposed data center campus in Decatur Township, Indianapolis, Indiana.

Planned for a roughly 130-acre site at 6400 Kentucky Avenue, the proposed campus would include two data center buildings totaling approximately one million square feet. The development would also include a new electrical substation and transmission infrastructure, with Sabey paying 100 percent of allocated costs so the project does not create additional costs for local ratepayers.

The project represents approximately $4.25 billion in projected taxable capital investment. Construction is expected to begin in the second quarter of 2027, with phased occupancy anticipated from the fourth quarter of 2029 through 2035. During the abatement period, the campus is estimated to rank as Marion County’s ninth-highest taxpayer, rising to the second-highest taxpayer after the abatement ends.

Sabey has also outlined several community commitments connected to the project. These include $5 million for updated roads in Decatur Township and nearly $25 million for a new aquatic center. The project is projected to contribute $21.9 million to Decatur Township public schools during the abatement period and $221 million over 30 years. The Marion County Library is projected to receive $2.2 million during the abatement period and $22 million over 30 years.

At full employment, the campus is projected to create and maintain 75 full-time permanent positions with average annual wages of at least $104,000 or 125 percent of Marion County’s average annual wage, whichever is higher at the time of job creation. Sabey also plans to use local union labor to construct and maintain the project.

“Indianapolis is a strategic addition to Sabey’s national platform, with the infrastructure, skilled workforce and business environment needed to support a long-term investment of this scale,” said Clete Casper, Director of Real Estate at Sabey Data Centers. “We are approaching this project not only as a developer and operator, but as a future neighbor with a responsibility to invest in the community alongside our facilities.”

The post Building for the Future: Sabey Data Centers Outlines Community Commitments for Proposed Indianapolis Campus appeared first on Data Center POST.

TL;DR

• Growing footprint: Empire Fiber Internet expanded its fiber network in Ithaca, NY, adding over 700 homes and businesses, including Collegetown, Bryant Park, and Belle Sherman.
• Milestone reached: The expansion brings Ithaca homes and businesses passed to more than 5,400
• Community backing: The Tompkins Chamber welcomed the investment as a boost to reliable internet options across the county.

# # #

Empire Fiber Internet, a leading fiber optic internet service provider serving communities across New York and Pennsylvania, has expanded its network in Ithaca, NY, making service available to more than 700 additional homes and businesses. The move builds on the company’s growing presence in the city, bringing the total number of homes and businesses passed to more than 5,400.

“Greater access to reliable, high-speed internet is important for Ithaca’s residents, students, non-profits, entrepreneurs and local businesses,” said Peggy Coleman, President & CEO of the Tompkins Chamber. “We’re pleased to see Empire Fiber Internet continue investing in the Ithaca community, adding another reliable service option within Tompkins County.”

The expansion delivers symmetrical upload and download speeds to support the demands of modern connected life, from remote work and online learning to streaming, gaming, and smart home and business technology. Customers are backed by a locally staffed, 24/7/365 support team, and the company continues to deepen its ties to the region as an active member of the Tompkins Chamber.

“Ithaca deserves internet that works as hard as the people who live and work here,” said Kevin Dickens, CEO of Empire Fiber Internet. “From students and families relying on connectivity for school and work to businesses that depend on our internet to serve customers, fiber delivers the speed and reliability that modern life requires. And just as important, our customers can count on responsive, local support when they need it.”

To learn more about Empire Fiber Internet, visit empirefiber.com.

The post Empire Fiber Internet Announces Service Expansion in Ithaca, NY appeared first on Data Center POST.

TL;DR

• New CFO: Christopher “Chris” DeAlmeida joined Duos as Chief Financial Officer, effective August 24, 2026.
• Financial strategy: He will oversee capital allocation, financial planning, and investment strategy as Duos scales its Edge AI platform.
• Industry experience: DeAlmeida brings more than 20 years of public-company finance and infrastructure leadership experience.

# # #

Duos Technologies Group, Inc. (“Duos” or the “Company”) (Nasdaq: DUOT) has appointed Christopher “Chris” DeAlmeida as Chief Financial Officer, effective August 24, 2026. DeAlmeida succeeds retiring CFO Adrian Goldfarb and joins Duos as the Company continues to scale its Edge AI platform and technology infrastructure solutions business.

“Chris is an experienced public company operator with a direct background in building and deploying capital for construction and infrastructure projects, making him a great fit for our business today and where we’re headed in the future,” said Doug Recker, CEO of Duos. “As we continue to transform the Company into a scaled operator within the broader AI ecosystem, we believe a proven financial executive like Chris presents an ideal combination of industry experience and financial acumen to lead this next stage in our evolution. On behalf of the board of directors and the rest of our management team, I’d like to formally welcome Chris to Duos.”

As CFO, DeAlmeida will oversee capital allocation, financial planning, and investment strategy as Duos advances its AI infrastructure growth plans. He brings more than 20 years of experience across public-company finance, capital markets, investor relations, mergers and acquisitions, and infrastructure development.

“The Duos team has made tremendous progress in building an AI infrastructure business in an incredibly short amount of time. They have demonstrated a clear ability to identify attractive market opportunities and successfully compete in high-demand environments. I look forward to joining the team at such an exciting time,” said Chris DeAlmeida. “Looking ahead, I plan to leverage my public company finance experience and extensive background in infrastructure development to strengthen our financial strategy and position Duos to capitalize on the significant opportunities ahead. I am confident in the team, our strategy, and our ability to create long-term value for our shareholders as we continue building the foundation for Duos’ next phase of growth.”

His previous leadership roles include CFO positions at Wrap Technologies, Inc. and Encore, as well as 11 years with Orion Group Holdings, Inc., where he served as CFO, Treasurer, and Executive Vice President. Across his career, DeAlmeida has completed more than 15 buy-side acquisitions and served as a CFO for founder-led and private equity-backed companies.

The appointment reflects Duos’ continued focus on strengthening its financial leadership as the Company scales its Edge AI platform and advances its long-term growth strategy.

To learn more about Duos Technologies Group, visit duostechnologies.com.

The post Duos Technologies Group, Inc. Names Christopher DeAlmeida Chief Financial Officer appeared first on Data Center POST.

TL;DR

• In January 2026, the first operational orbital data center nodes were launched into low Earth orbit by Axiom Space and Kepler Communications, transitioning space-based computing from a theoretical concept to physical, working infrastructure.
• Moving computing into orbit provides continuous, unfiltered solar power, avoids Earth-based competition for land, water, and power grids, and allows local processing of satellite-collected data to significantly reduce downlink transmission volumes.
• Massive technical challenges remain, particularly heat dissipation in a vacuum (relying solely on radiation), hardware vulnerability to cosmic radiation, high launch costs, and the complexity of in-orbit maintenance.
• The global space-based orbital data center market is projected to grow from $1.2 billion in 2025 to $7.8 billion by 2034 (a 22.5% CAGR), acting as a complementary, specialized layer to traditional terrestrial infrastructure

# # #

Data centers have always been built around three constraints: land, power, and cooling. For decades, the solution was simple: build bigger facilities closer to cheap electricity and cool water. But as artificial intelligence workloads push electricity demand to levels that are straining regional power grids, a small but growing group of aerospace companies, cloud providers, and research agencies are asking a different question: what if the data center didn’t need to be on Earth at all?

That question is no longer purely theoretical. In January 2026, the first operational orbital data center nodes reached low Earth orbit, marking a shift from concept studies to working infrastructure. This article looks at what orbital data centers actually are, how they work, who is building them, and the real engineering and economic hurdles that stand between today’s early prototypes and a genuine off-world computing industry.

What Is an Orbital Data Center?

An orbital data center is a satellite, or network of satellites, equipped with computing hardware, such as processors, storage, and networking equipment,  that performs data processing in space rather than on the ground. Instead of beaming raw satellite imagery, sensor readings, or telemetry back to Earth for analysis, an orbital data center can process that data locally, in orbit, and send back only the results.

The concept builds on a technique already common in terrestrial computing: edge processing. Just as a smart camera might analyze video on-device rather than streaming it all to a distant server, an orbital compute node can filter, compress, or run inference on data before it ever reaches a ground station. The difference is the “edge” here sits hundreds of kilometers above the planet.

Why Move Computing Into Orbit?

A handful of physical realities make space an unusual, but not entirely irrational, place to put servers.

• Continuous, unfiltered sunlight. Outside Earth’s atmosphere, solar panels aren’t diminished by clouds, dust, or nighttime, and in the right orbit they can receive sunlight far more intensely and consistently than panels on the ground. Satellites in specific low-Earth orbits can be positioned to stay in near-constant daylight, reducing the need for heavy battery storage.
• No land, water, or grid competition. Ground-based data centers increasingly compete with homes and industry for electricity, land, and municipal water used in cooling. In orbit, none of those resources are contested. Though, as discussed below, cooling in space brings its own very different problems.
• Proximity to the data source. A growing share of the data that needs processing, like Earth observation imagery, weather data, communications traffic, already originates from satellites. Processing it in orbit, near where it’s collected, can cut down on the volume of raw data that needs to be transmitted to the ground, easing pressure on radio spectrum and ground station capacity.
• Radiation-hardened isolation. Some research and defense applications benefit from computing environments that are physically separated from terrestrial networks, which can offer security and sovereignty advantages for sensitive workloads.

How the Technology Actually Works

A typical orbital data center architecture involves a few connected pieces:

1. Compute payloads: Satellites or modules carrying processors (increasingly GPUs or AI accelerators such as tensor processing units) capable of running machine learning inference, data fusion, or general cloud computing tasks.
2. Optical inter-satellite links: Instead of routing data through radio-frequency ground stations, many orbital systems use laser-based optical communication between satellites, which can move data at very high speeds without waiting for a satellite to pass over a ground antenna.
3. Solar arrays and thermal systems: Large solar panels supply power, while radiators dissipate heat. Because there is no air in space to carry heat away by convection, spacecraft must shed heat through radiation alone, which is one of the toughest engineering constraints on orbital computing.
4. Radiation-tolerant hardware: Space is a harsher environment for electronics than Earth’s surface. Cosmic rays and solar radiation can corrupt data or damage chips over time, so components need shielding, error correction, or radiation-hardened design.

Who Is Building Them?

Interest in orbital computing has broadened considerably over the past two years, moving from single research proposals to a genuine multi-organization field.

• Axiom Space and Kepler Communications deployed some of the first operational orbital data center nodes in January 2026, riding on an optical relay satellite constellation designed for inter-satellite mesh networking. The hardware grew out of an earlier data-processing prototype that had already been tested aboard the International Space Station.
• Google’s Project Suncatcher is a research program exploring whether tightly clustered constellations of solar-powered satellites, carrying the company’s own AI accelerator chips and linked by high-bandwidth optical connections, could eventually scale machine learning compute in orbit. Google has described early radiation testing of its hardware and plans prototype satellite launches with a partner in the following years.
• Starcloud, a startup backed by chipmaker investment, is developing orbital computing platforms built around commercial GPU hardware, with ambitions to scale toward much larger orbital compute capacity over time.
• The European Commission, through its Horizon Europe research program, has funded feasibility studies examining whether orbital data centers could support Europe’s digital sovereignty and climate goals.
• Several other aerospace, cloud, and satellite communications companies have filed plans or committed funding toward orbital compute projects, and industry analysts increasingly track “in-orbit data centers” as a distinct market category.

The Engineering Challenges

Despite the momentum, orbital data centers face several unresolved technical problems.

1. Heat dissipation. On Earth, data centers use fans, liquid cooling, and outside air to remove heat generated by densely packed servers. In space, with no atmosphere to carry heat away, spacecraft rely entirely on radiating heat into the vacuum, which requires large radiator surfaces relative to the amount of computing power onboard. This is widely considered one of the hardest constraints on how much compute density an orbital system can realistically support.
2. Radiation and hardware reliability. Electronics in orbit are exposed to cosmic radiation that can flip bits, corrupt memory, or gradually degrade components. Systems either need radiation-hardened chips, which tend to lag behind the newest commercial processors in raw performance, or software-level error correction and redundancy to compensate for using standard commercial hardware.
3. Maintenance and servicing. Ground-based data centers benefit from technicians who can swap failed drives or replace servers within minutes. Servicing distributed hardware in orbit is far more complex and costly, meaning orbital systems generally need to be designed for long, largely unattended operation, with failure built into the calculus from the start.
4. Launch costs and scale. Even with falling launch prices, getting substantial computing hardware into orbit remains expensive relative to building on the ground. Achieving the kind of gigawatt-scale computing capacity that some proposals envision would require assembling very large structures in orbit, like solar arrays and radiators potentially spanning kilometers, which is well beyond what any single rocket can currently launch in one piece.
5. Space traffic and orbital debris. A larger population of satellites in low Earth orbit adds to congestion in an already crowded orbital environment, raising collision risk and the amount of tracked debris. Astronomers have also raised concerns that dense satellite constellations can interfere with ground-based telescope observations.

How Orbital Compute Might Fit Alongside Ground-Based Data Centers

Most current proposals don’t describe orbital facilities as a wholesale replacement for terrestrial data centers. Instead, they tend to fall into a few likely near-term roles:

• Pre-processing satellite data before it’s sent to Earth, reducing the volume of raw data that needs to be downlinked.
• Supporting AI inference for space-based applications, such as autonomous satellite operations or rapid analysis of Earth observation imagery.
• Offering an additional, geographically and energy-independent compute layer for organizations that want workloads physically separated from terrestrial infrastructure.

Large-scale AI model training, the same kind of workload driving today’s biggest data center investments on Earth, remains a much more distant prospect in orbit, given the current limits on power generation, heat rejection, and hardware scale achievable in space.

Market Outlook

According to Research Intelo, the global space-based orbital data center market was valued at $1.2 billion in 2025 and is projected to reach $7.8 billion by 2034, expanding at a compound annual growth rate (CAGR) of 22.5% over the forecast period 2026-2034, driven by rapid advances in in-orbit computing infrastructure, proliferating low Earth orbit (LEO) satellite constellations, and the urgent demand for real-time AI in space exploration applications that require latency-sensitive processing far from terrestrial networks.

Terrestrial data center investment, meanwhile, continues to expand rapidly and shows no sign of slowing because of orbital experimentation. Analysts generally view orbital and ground-based computing as complementary rather than competing paths for the foreseeable future, with space-based systems most likely to influence the economics of remote or extreme-environment terrestrial sites, such as Arctic or underwater data centers, before they meaningfully compete with mainstream cloud infrastructure.

The Bottom Line

Orbital data centers have moved from a speculative idea to a small but real category of operating infrastructure, backed by aerospace companies, major cloud providers, and government research agencies. The core appeal, which includes abundant solar power, no land or water competition, and proximity to satellite-generated data, is genuine. So are the obstacles: heat dissipation in a vacuum, radiation-hardened hardware, in-orbit servicing, and the sheer cost of launching and assembling meaningful compute capacity in space.

For now, orbital computing is best understood as an emerging, specialized layer of the broader data infrastructure landscape. It is worth watching closely, but it remains years away from reshaping how most of the world’s computing gets done.

# # #

About the Author

Mandar Shewale is a Marketing Specialist at Research Intelo with experience in digital marketing, SEO, content planning, and market research communications. He works on content development, search optimization, and marketing coordination across research-focused projects. His work involves supporting online visibility initiatives and maintaining content quality across digital platforms.

The post Space-Based Orbital Data Centers: How Computing Is Moving Beyond Earth appeared first on Data Center POST.

TL;DR

• Recent shifts in VMware’s licensing and commercial model are forcing IT leaders to reevaluate virtualization strategies. DāSTOR provides a secure enterprise private cloud to keep existing workloads running smoothly, eliminating the need to rush critical decisions.
• DāSTOR supports a Bring Your Own License (BYOL) approach, allowing organizations to preserve their current VMware investments and continue utilizing the familiar tools and management processes their IT teams already know.
• Replacing VMware overnight is risky and complex. Utilizing a hosted private cloud allows businesses to support business-critical workloads seamlessly while giving them the time to evaluate future virtualization platforms on their own terms.
• Rather than being forced into immediate action by an upcoming license renewal, hosting VMware workloads gives organizations the breathing room to align future technology investments with long-term business goals.

# # #

Written by iMiller Public Relations on behalf of DāSTOR.

Recent changes to VMware’s licensing and commercial model are prompting IT leaders to carefully reevaluate their virtualization and infrastructure strategies. Because many organizations rely on VMware to support business-critical workloads, rushing into an immediate platform replacement is highly impractical, complex, and risky. Instead of executing an abrupt migration under external pressure, organizations are looking for ways to maintain operational stability while preserving the flexibility to evolve on their own terms.

To address this need, DāSTOR offers a secure enterprise private cloud platform that hosts and supports existing VMware workloads. This solution features a Bring Your Own License (BYOL) approach, allowing businesses to use their current VMware licenses to preserve prior technology investments and continue leveraging the familiar tools and management processes their IT teams already know. By hosting VMware workloads in this secure environment, organizations can bypass the pressure of upcoming licensing renewals or hardware refreshes and instead make deliberate technology choices on their own timeline.

Ultimately, maintaining a stable hosted foundation gives IT teams the breathing room to build a roadmap aligned with long-term business goals, whether they choose to run VMware for years or gradually transition to new platforms as needs change. This measured approach minimizes migration risk and ensures that today’s infrastructure decisions do not restrict tomorrow’s opportunities. By providing the necessary security, time, and control, DāSTOR serves as a stabilizing partner to help organizations navigate these industry shifts with confidence.

This content originally appeared on DāSTOR’s website and has been adapted for syndication on Data Center POST. Read the complete article here.

The post Navigating VMware Change Without Rushing the Next Decision appeared first on Data Center POST.

TL;DR

• AI infrastructure growth is placing greater pressure on power availability, site selection, construction timelines, and operational readiness.
• High-density facilities require power, cooling, compute, and network systems to be designed as an integrated architecture.
• Successful deployment depends on disciplined execution across construction, commissioning, supply chains, and operational handoffs.
• Community engagement and responsible resource management are becoming essential to long-term data center development.

# # #

Artificial intelligence may be driving unprecedented demand for digital infrastructure, but turning that demand into operational capacity remains a complex challenge. That reality shaped discussions at the Data Center Frontier Trends Summit 2026, held August 4-6 in Reston, Virginia, where industry leaders examined what it takes to build, power, scale, and operate next-generation data centers.

Across three days of sessions, roundtables, networking, and data center tours, the event moved beyond industry projections to focus on execution. Conversations explored how power constraints, rising rack densities, compressed construction schedules, supply chain risks, and community concerns are influencing where projects are developed and whether they ultimately reach operation.

Power and Site Readiness Are Defining Development Strategies

Power availability has become one of the most important factors determining where data center projects can move forward. With utility interconnection delays and grid congestion affecting established markets, developers are increasingly considering behind-the-meter generation, battery storage, microgrids, and other alternative energy strategies.

During the opening keynote fireside chat, Lee Kestler, CEO of EdgeCore Digital Infrastructure, joined Data Center Frontier Founder Rich Miller and Editor-in-Chief Matthew Vincent to examine how power, entitled land, and execution speed are reshaping the geography of AI infrastructure. The conversation addressed how developers are selecting markets, working with utilities and governments, and preparing for continued hyperscale and enterprise AI growth.

Site selection now extends well beyond securing land in a desirable market. Developers must evaluate power, fiber, water, permitting requirements, and community acceptance together, recognizing that a site that works in one area may present a significant constraint in another.

AI Factories Move From Design to Operation

The industry has become increasingly sophisticated in designing AI environments through simulation, digital twins, and GPU-accelerated modeling. However, operating ultra-high-density infrastructure introduces challenges that may not become apparent until equipment is installed, integrated, commissioned, and placed into production.

Steve Altizer, CEO and President of Compu Dynamics, participated in “The AI Factory in Practice: Designing and Operating at Scale.” The panel examined the operational issues that can emerge when AI workloads go live, including liquid cooling integration, evolving power topologies, commissioning risks, infrastructure orchestration, and the transition from construction to day-to-day operation.

The discussion reflected an important shift in how AI data centers are evaluated. Technical design remains essential, but competitive advantage increasingly depends on whether an organization can successfully integrate and operate complex infrastructure at scale.

Construction Timelines Increase Pressure on Project Execution

AI infrastructure projects are growing larger while delivery schedules are becoming more compressed. Labor availability, long-lead equipment, site readiness, modular construction strategies, and supply chain volatility can all affect whether announced capacity is completed on schedule.

Kurt Haglund, Chief Operating Officer of Compu Dynamics, joined an interactive roundtable focused on AI data center density, design, and deployment. The discussion explored real-world deployment lessons, emerging design patterns, and the practical constraints affecting next-generation infrastructure as operators work to support higher rack densities and more demanding power and network requirements.

These challenges continue through the final stages of development when generators, switchgear, UPS systems, controls, fuel systems, utility coordination, and commissioning activities must come together. Even projects that have secured land, power, financing, and equipment can face delays if late-stage testing, integration, and operational handoffs are not carefully coordinated.

Cooling Strategies Must Support a Range of AI Workloads

AI infrastructure does not have a single density or cooling requirement. Large-scale training environments may require purpose-built facilities and extensive liquid cooling, while enterprise inference deployments may operate at lower densities or within existing data center environments.

Lindsay Schulz, Global Principal Technologist at Equinix, participated in the closing day panel, “From Hybrid to Liquid: Scaling Cooling Infrastructure for the Next Era of AI.” The session addressed direct-to-chip cooling, facility retrofits, hybrid air-and-liquid environments, heat rejection, water stewardship, and the operational preparation needed to support sustained high-density deployments.

The transition to liquid cooling involves more than installing new thermal management equipment. Operators must also consider serviceability, facility integration, workforce training, maintenance procedures, and the long-term resiliency of increasingly complex cooling architectures.

Community Acceptance Becomes a Development Requirement

As data centers consume more power, land, and water, proposed developments are receiving greater attention from policymakers and surrounding communities. Permitting timelines and long-term project viability increasingly depend on whether developers can demonstrate responsible resource use, transparency, and meaningful local economic benefits.

Buddy Rizer, Executive Director of Loudoun County Economic Development, joined the closing keynote examining stewardship, sustainability, and the industry’s social license to operate. The discussion focused on how siting, construction practices, workforce development, local participation, and community engagement influence trust as AI-led infrastructure expands.

Community acceptance cannot be addressed solely through messaging after a project has been announced. It is shaped by decisions made throughout site selection, design, construction, and operation, making early engagement an increasingly important part of successful data center development.

Turning Infrastructure Plans Into Operational Capacity

The Data Center Frontier Trends Summit 2026 highlighted the widening difference between announcing an AI infrastructure project and successfully bringing it online. Demand and capital remain strong, but power, equipment, labor, permitting, commissioning, and operational readiness ultimately determine which projects reach production.

The next phase of data center growth will require close coordination across developers, utilities, construction firms, technology providers, operators, investors, and communities. Organizations that integrate these disciplines early will be better positioned to translate AI infrastructure plans into reliable, scalable capacity.

For additional insights and information about upcoming events, visit the Data Center Frontier Trends Summit website, www.dcftrends.com.

The post From AI Ambition to Infrastructure Execution: Key Trends Shaping Data Center Development appeared first on Data Center POST.

Company Profile: Ilkari’s Vision for a More Resilient, Open Future

Data Center POST recently connected with Shane Paterson, the Founder and CEO of Ilkari. With almost three decades of leadership experience across six countries, he brings a global perspective and extensive expertise to Ilkari’s mission: enabling enterprises to take control of their sensitive data or workloads.

Passionate about tech and not afraid to challenge the status quo, Shane’s background spans IT services, cybersecurity and data centre management, with senior roles at some of the world’s leading technology providers. Starting out as an engineer, he has built a strong track record of delivering complex, high-impact projects throughout Europe, the Nordics, South America, Australia, and South East Asia.

At Ilkari, he combines his technical acumen and strategic leadership to help organizations meet data legislation requirements, reduce operational costs and scale more effectively.

What does Ilkari do?

Ilkari is a sovereign infrastructure provider delivering data centre services, sovereign cloud, and digital infrastructure solutions across Europe, and Latin America.

The company operates certified, high-performance infrastructure environments designed for operational resilience, regulatory alignment, and long-term digital sovereignty, supporting enterprise, regulated-industry, and data-intensive workloads.

Ilkari’s sovereign infrastructure platform spans data centre operations in Iceland, Croatia, and Colombia. Ilkari also operates a premium domain management and digital intelligence platform.

What problems does Ilkari solve in the market?

As a sovereign infrastructure provider delivering data centre services, sovereign cloud, and digital infrastructure solutions across Europe and Latin America, Ilkari helps companies solve the need for secure, compliant, regionally controlled digital infrastructure for enterprises and regulated industries.

What are Ilkari’s core products or services?

llkari operates across three closely connected areas:

1. Certified data centre and colocation services, including carrier-neutral connectivity and 24/7 onsite technical support.
2. Sovereign cloud infrastructure, delivered through shared and dedicated deployment models.
3. Corporate domain name services covering the registration, management and protection of digital identities globally.

Together, these services allow organizations to manage a broader part of their infrastructure through one accountable provider: from their domain portfolio and digital identity to the physical and cloud environments supporting their applications and data.

Our data centre footprint currently spans Colombia, Iceland, and Croatia. This gives customers access to infrastructure across Latin America and Europe, supported by local operational teams and international expertise.

What markets do you serve?

We primarily serve medium-sized and large organizations operating business-critical, regulated or data-intensive environments. These include financial services, government, and public-sector organizations, telcos, healthcare, and technology, as well as digital-first businesses and online platforms managing complex international domain portfolios.

Geographically, our infrastructure platform serves customers across Europe and Latin America, with data centre operations in Croatia, Iceland, and Colombia. Our domain name services have a broader international reach, supporting organizations in managing and protecting their or their customers’ digital identities across global markets.

We are particularly relevant to organizations that need to keep sensitive workloads within a defined jurisdiction, diversify infrastructure across regions or providers, establish business continuity outside a congested primary market, or reassess their dependence on hyperscale cloud and virtualisation vendors.

We also support European organizations seeking to grow or invest in Colombia, and Latin American businesses looking for infrastructure and expertise that can support their international expansion. Our combination of local teams with regional expertise and cross-regional operations allows us to serve as the link between those markets.

What challenges does the global digital infrastructure industry face today?

The industry is facing a collision between rapidly growing demand and increasingly constrained supply.

AI is accelerating infrastructure requirements, but it is arriving at a time when many established data-centre markets are already approaching capacity limits. Suitable land is scarce, grid connections can take years and power constraints are delaying or in some locations effectively freezing further development. According to the IEA, global data-centre electricity consumption increased by 17% in 2025 and is expected to double by 2030, with AI-focused facilities growing considerably faster. That makes access to reliable power one of the industry’s defining constraints.

This growth also has to be sustainable for the communities hosting it. Data centres compete with households and other industries for energy, land, and water. Operators therefore need to demonstrate tangible local value, invest in efficiency and renewable energy, and engage transparently with communities. The industry cannot assume that technical demand automatically creates permission to build.

A second challenge is infrastructure concentration. organizations have gained enormous flexibility from cloud, but many now depend heavily on a small number of hyperscalers, software vendors, and geographic regions. Recent changes in virtualisation licensing and pricing have made this dependency tangible: technology choices made several years ago can suddenly become major cost and operational constraints. For an IT leader, migrating hundreds of virtual machines, applications and data repositories is not a quick procurement decision. Lock-in reduces negotiating power and can make the path to an alternative complex and expensive.

Regulation is adding further pressure. Frameworks such as the EU Digital Operational Resilience Act (DORA) require organizations to look beyond basic compliance and consider operational resilience, ICT third-party risk and concentration across critical providers. These are no longer questions that can be delegated entirely to the IT department; they increasingly affect enterprise risk and board-level decision-making.

This is also why sovereignty is moving into mainstream infrastructure procurement. PwC found that 44% of end users now rank sovereignty and data residency as a leading buying priority; almost level with the 46% prioritising total cost of compute. That tells us customers are no longer evaluating infrastructure on price and performance alone. They also want to know where their workloads reside, which jurisdiction governs them, who can access them and whether they have a credible exit path if their requirements change.

The central challenge is therefore not simply building more capacity. It is building infrastructure that is powerful enough for the next generation of workloads, sustainable enough to retain community support and sufficiently open, resilient and accountable to give customers meaningful control.

How is Ilkari adapting to these challenges?

We are building a multi-region infrastructure platform around four principles: regional capacity with a focus on emerging markets, operational control, customer choice, and sustainability and community impact.

First, we are investing in locations that can offer credible alternatives to the most constrained digital infrastructure hubs. The integration of DC North in Croatia, for example, gives customers access to a fully operational 4 MW carrier-neutral data centre within the European Union, with direct connectivity to internet exchange points in Austria, Hungary, and Slovenia. It adds capacity in a strategically useful location while supporting infrastructure diversification across Central and Eastern Europe.

Second, we combine certified data centres and colocation with a shared and dedicated sovereign cloud. This gives customers more flexibility in deciding where individual workloads should run and how much isolation, control and support they require. A regulated workload may require dedicated infrastructure within a defined jurisdiction, while another application may be better suited to a shared environment. Sovereignty should provide practical choices rather than impose a single architecture.

Third, we recognise that many organizations want to reduce platform dependency but cannot simply redesign their entire technology estate. Our role is to provide a credible alternative and a realistic migration path, supported by local engineers and clear accountability. That is particularly relevant for businesses reassessing virtualisation costs, looking to repatriate selected workloads from hyperscale cloud, or seeking a secondary environment to strengthen resilience.

AI reinforces this strategy. We are not positioning Ilkari as an AI company or applying an “AI-ready” label indiscriminately. Our focus is on the foundations that AI and other data-intensive workloads require: dependable power, resilient facilities, strong connectivity, predictable performance and clear governance over the underlying data.

Fourth, we embed sustainability and community impact into our infrastructure decisions. In Colombia, our facilities operate on a national grid powered predominantly by renewable energy, with hydropower accounting for the majority of generation, and are certified under the SS 564 Green Data Centre standard. In Croatia, DC North achieves a partial PUE (pPUE) of ≤ 1.14, holds an A+ energy-efficiency certificate and is ISO 14001 certified for environmental management. In Iceland, our infrastructure is powered by 100% renewable energy from a mix of hydroelectric power, fed by glacial rivers and melting ice, and geothermal energy from volcanic heat. Across our offices, we run frequent ESG initiatives, including local tree-planting activities, giving back to nature and supporting local communities.

Ultimately, customers should not have to choose between modern infrastructure and meaningful control. Our objective is to give them more choice over where their workloads operate, how their data is governed and which dependencies they are prepared to accept.

What are your company’s key differentiators?

Our principal differentiator is that we approach sovereignty across the infrastructure stack rather than presenting it as a feature attached to one product.

We combine corporate domain name services, certified data centres, colocation, and sovereign cloud. This gives customers a clearer line of accountability, from their digital identity through to the physical and cloud infrastructure supporting their applications and data.

Our geographic strategy is also deliberate. We are not attempting to replicate the scale or footprint of a hyperscaler. We invest selectively in markets where our infrastructure can offer meaningful capacity, connectivity and jurisdictional choice. Croatia, Iceland, and Colombia each provide distinct strategic advantages across Europe and Latin America.

We also combine international expertise with local operational teams. Sovereignty cannot be delivered through contractual language alone. It depends on who operates the infrastructure, who has access to it, who responds when support is required, and whether the customer can reach accountable technical experts.

Finally, we provide a credible alternative for organizations seeking to reduce dependency on large cloud and virtualisation vendors. That does not mean every workload must leave the hyperscalers. It means customers should have genuine options, clearer economics, and a viable path to move or diversify workloads if their requirements change.

What can we expect to see/hear from Ilkari in the future?

Our immediate priority is to deepen the capabilities of the platform we have already established.

In Europe, that means integrating DC North into Ilkari’s wider operations and developing its role as a resilient infrastructure and interconnection hub for Croatia and the wider Central and Eastern European region. We see a clear opportunity to support organizations seeking greater infrastructure diversity and jurisdictional choice within the EU.

Across the business, we will continue developing our sovereign cloud capabilities and the services surrounding them, particularly in areas such as workload migration, security, connectivity, resilience and infrastructure management.

We also expect AI to increase demand for infrastructure that combines scalable capacity with clear jurisdiction, predictable performance and strong data governance. Ilkari will continue developing its data centre and cloud capabilities to support data-intensive workloads in markets where organizations require greater control over how and where their data is processed.

Over time, you can expect Ilkari to establish a presence in additional markets where regulation, data growth and infrastructure concentration are creating demand for credible alternatives. Our expansion will be selective. The objective is not to accumulate locations, but to build a connected platform in which each market serves a clear operational and customer purpose.

What upcoming industry events will you be attending?

Our next major ICT event will be ANDICOM in Cartagena at the beginning of September. Ilkari will exhibit at the event, contribute to its conference programme, and participate as part of the European Union delegation, with the EU joining this year as guest region.

We will showcase our sovereign cloud and certified data centre capabilities while sharing a practical European perspective on sovereign infrastructure with Colombian businesses and government organizations. As a company operating certified data centres across Europe and Colombia, we also have a valuable role to play as a local infrastructure partner for European organizations looking to grow or invest in the Colombian market.

Before then, we will attend Asobancaria’s Banking Convention in Cartagena, which brings together senior leaders from Colombia’s banking and financial-services sector.

Following the integration of DC North, we are expanding our engagement with data centre, cloud and wider ICT communities in Croatia and across Central and Eastern Europe. This will form an important part of our industry programme as we strengthen Ilkari’s regional presence and develop DC North’s role within the European digital infrastructure and ICT ecosystem.

Do you have any recent news you would like us to highlight?

Our most significant recent development is the integration of the data centre DC North in Varaždin, Croatia, into Ilkari’s European operations.

DC North is a fully operational, 4 MW carrier-neutral data centre located approximately 70 kilometres north of Zagreb. It is Croatia’s largest carrier-neutral data centre by operational power capacity and provides connectivity to internet exchange points in Austria, Hungary, and Slovenia.

The integration expands Ilkari’s operational capacity within the European Union and extends our data centre footprint to three markets: Croatia, Iceland, and Colombia. It strengthens our ability to support regulated, sovereignty-focused, and data-intensive organizations across Europe, while providing an additional location for infrastructure diversification and business continuity.

It also creates a foundation for bringing together colocation, connectivity, and sovereign cloud capabilities within an EU jurisdiction.

Earlier this year, we also expanded our infrastructure platform with Ilkari Cloud. Delivered through shared and dedicated deployment models, Ilkari Cloud is designed to provide predictable performance, operational control, and clearer jurisdiction over customer data and workloads. It offers organizations an alternative as they reassess hyperscaler concentration, virtualisation costs and long-term platform dependencies.

Is there anything else you would like our readers to know about Ilkari and capabilities?

Sovereignty is sometimes presented as though organizations must choose between control and innovation. We do not believe that should be the case.

Businesses need infrastructure that performs, scales, and supports technologies such as AI. But they also need to understand the dependencies they are creating and retain meaningful control over their data, providers and operational choices.

Ilkari was established to help close that gap. We are building an alternative for organizations that want modern infrastructure without losing sight of jurisdiction, accountability, or resilience.

We are still a relatively young company, but the foundations are tangible: operational data centres across three markets, sovereign cloud capabilities, an international domain services business, and local teams with substantial infrastructure experience.

Our ambition is straightforward: to give organizations greater control over the digital infrastructure on which they increasingly depend.

Where can our readers learn more about Ilkari?

Readers can learn more about us on our website, ilkari.tech/about. You can also contact us through the contact form on our website.

# # #

About Data Center POST

Data Center POST provides a comprehensive view of the digital infrastructure landscape, delivering industry insights into the global data center ecosystem. As the industry’s only peer-contributed and online publication, we offer relevant information from developers, managers, providers, investors, and trendsetters worldwide.

Data Center POST works hard to get the most current information and thought-provoking ideas most apt to add relevance to the success of the data center industry. Stay informed, visit www.datacenterpost.com.

If you are interested in contributing to Data Center POST, contact us at contributions@datacenterpost.com or submit your article here.

Want more digital infrastructure news? Stay in the know and subscribe to Data Center POST today!

The post Company Profile: Ilkari’s Vision for a More Resilient, Open Future appeared first on Data Center POST.

TL;DR

• Evaluating cooling components (such as cold plates) in isolation can be misleading; thermal and hydraulic performance must be engineered holistically to prevent degrading overall system efficiency.
• As liquid cooling scales across thousands of servers, small flow resistance inefficiencies accumulate, increasing pump power demands and potentially limiting infrastructure scaling.
• Much like Formula One teams constantly refine thermal packages for each race, data center cooling designs must evolve quickly to support the rapid arrival of new GPU generations with distinct thermal profiles.
• Computational fluid dynamics (CFD) and additive manufacturing (3D printing) allow engineers to design complex, monolithic cold plates that optimize both heat transfer and pressure drop while eliminating potential leak paths.

# # #

As cooling efficiency is emerging as a key constraint in scaling AI data center infrastructure, Michael Fuller, Executive Chairman of Conflux Technology and former F1 engineer, explores how Formula One’s system-level approach to thermal management could help data centers balance cooling performance with energy efficiency and operational demands.

As AI infrastructure scales, most conversations around data center performance still focus on compute capacity: faster GPUs, denser racks and increasing power availability. However, a critical constraint is quietly emerging alongside these advances – cooling efficiency.

While the industry has largely acknowledged that liquid cooling is necessary to support next-generation hardware, less attention has been paid to how efficiently those systems operate at scale. That distinction – between simply cooling hardware and engineering an efficient thermal system – is one Formula One has been refining for decades.

Engineering That Never Stands Still

Perhaps Formula One’s greatest lesson isn’t simply how to engineer an efficient cooling system, but how to engineer one that can evolve. Every race weekend brings new components, revised aerodynamic packages and evolving thermal demands, leaving engineers with little time to redesign, test and validate solutions. That relentless cycle has fostered an engineering mindset built around rapid iteration, where cooling systems are expected to evolve alongside the hardware they support. AI data centers are now entering a similar phase. Successive generations of GPUs are arriving faster than ever, each with different thermal profiles, power densities and packaging constraints, meaning cooling solutions can no longer remain static while the technology around them advances.

From Thermal Capacity to System Efficiency

That evolution also changes the engineering challenge. It is no longer enough for cooling systems to remove the required amount of heat. They must do so with maximum efficiency.

AI workloads are pushing rack power densities to unprecedented levels, and simply meeting thermal requirements does not guarantee an optimal system. Instead, operators must consider how much energy is required to deliver that cooling capacity.

This shifts the focus from peak performance to system efficiency, a change that has significant implications for both operational cost and infrastructure design.

The Hidden Cost of Pressure Drop

One of the most important, and often overlooked, aspects of cooling performance is pressure drop.

In liquid cooling systems, pressure drop has a significant influence on the energy required to move coolant through the system. While small, perhaps even negligible, at the component level, its impact becomes significant when multiplied across thousands of servers.

Inefficient flow paths can increase pump power requirements, raise energy consumption and reduce overall system flexibility. In extreme cases, they can become a limiting factor in scaling infrastructure.

Why Component-Level Thinking is No Longer Enough

Components such as cold plates are often assessed in isolation, focusing on metrics like heat transfer performance and flow rate. However, as systems become more complex, this approach can be misleading. Formula One engineers have long understood that a component delivering impressive standalone performance can still compromise the performance of the car as a whole.

A component that delivers strong thermal performance but induces high hydraulic resistance may degrade overall system efficiency. The net result is higher energy use, even if individual components appear optimised. To address this, cooling must be treated as a system-level engineering challenge, where thermal and hydraulic performance are considered together rather than separately.

Balancing Competing Requirements

Effective liquid cooling requires managing a fundamental trade-off between improving heat transfer and maintaining low resistance to flow. Design approaches that increase turbulence or surface area can enhance thermal performance, but they often do so at the cost of higher pressure drop. The challenge lies in achieving both low thermal resistance and low hydraulic resistance at the same time.

This balance demands precision engineering at both macro and micro levels. The most effective solutions focus on directing coolant exactly where it is needed, rather than applying uniform cooling across an entire component.

At Conflux, for example, our latest single-phase cold plate designs achieve thermal resistance as low as 0.0076°C/W at 6 LPM across a 2,500 mm² footprint, while maintaining exceptionally low pressure drop – demonstrating that high thermal performance and hydraulic efficiency can be achieved together, rather than traded off against one another. It’s the same engineering philosophy that underpins Formula One: the best solution isn’t the one that excels in a single metric, but the one that delivers the greatest overall system performance.

The Formula One Engineering Mindset

One of Formula One’s defining characteristics is that engineers do not evaluate components in isolation. Every design decision is based on its impact on overall system performance. A component that performs exceptionally well in isolation can still reduce the overall performance of the car if it compromises weight, reliability or aerodynamic efficiency.

This principle is directly relevant to data centers, where cooling systems must balance competing objectives under tight constraints. This is why cooling design is increasingly moving beyond component optimisation towards system-level engineering, where every design decision is judged by its impact on overall performance.

Delivering that level of optimisation demands equally sophisticated design tools and manufacturing techniques.

The Role of Advanced Design and Manufacturing

Achieving high levels of efficiency in liquid cooling begins with how fluid flow is controlled within components.

Modern engineering approaches use simulation tools such as computational fluid dynamics to analyse and refine internal geometries. These tools make it possible to optimise coolant distribution, minimise losses and improve heat transfer where it matters most.

It’s an approach familiar to Formula One, where advanced simulation and rapid manufacturing allow engineers to continually refine components in pursuit of marginal gains.

In liquid cooling, those same principles are enabling a new generation of highly optimised cold plate designs.

Additive manufacturing makes this possible by enabling complex internal channels and geometries that would be difficult or impossible to produce using conventional manufacturing methods. This gives engineers far greater control over coolant flow, allowing thermal performance and pressure drop to be optimised together rather than traded against one another.

It also enables cold plates to be produced as single, monolithic components rather than assemblies of multiple parts. By eliminating bonded interfaces, potential leak paths are eliminated while improving structural integrity – a significant advantage in mission-critical data center environments, where reliability is just as important as performance.

Rethinking the Role of Cooling

Formula One has always been a sport of marginal gains, where countless small engineering improvements combine to create a meaningful competitive advantage. AI data centers are entering a similar era.

As processor power continues to increase, the question is no longer simply whether cooling systems can remove enough heat. It is whether they can do so with maximum efficiency, balancing thermal performance, pressure drop, energy consumption and reliability across the entire system.

As a result, the next phase of data center innovation may be defined not only by faster processors, but by smarter and more efficient ways of keeping them cool.

# # #

About the Author

Michael Fuller’s career was launched off the back of his boyhood dream of designing racing cars. In 1999 he migrated to Europe as a Design Engineer, and over 15 years he held various design engineering and management roles in motorsport organisations competing in series as varied as the World Endurance Championship, World Rally Championship, and the Formula 1 World Championship. As an early adapter of Additive Manufacturing (AM) in the motorsport industry he was acutely aware of the opportunities available to enhance performance.

After returning to Australia, he pursued a design idea on an AM heat exchanger that proved to have ground-breaking performance potential and launched Conflux soon thereafter. Conflux has now grown into a world leading organisation that combines specialisations in heat transfer and additive manufacturing.

He is actively involved in all aspects of the business with significant effort in design engineering, relationship building, partnership research projects and industry speaking engagements. A considerable amount of his time is invested in strategy development and ensuring the right resources are in place to fulfil our vision.

The post Why AI Data Centers Should Think More Like Formula One Teams appeared first on Data Center POST.

TL;DR

• Data centers are so vital to modern life that the UK has granted them critical national infrastructure status, elevating data and energy security to matters of national security.
• While AI-driven energy demand is rising rapidly, the real hurdle is managing highly unpredictable, variable daily load patterns rather than just generating more total power.
• Resilience requires a flexible electricity system capable of balancing supply and demand in real time using solutions like energy storage, dispatchable generation, smart energy management, and workload shifting.
• Energy flexibility has become a business differentiator with companies like Google and Meta diversifying their power sources, while grid trials and rapid battery storage growth show real momentum.

# # #

Data Centers are so essential to modern life that, in 2024, the UK granted them the same critical national infrastructure status as water, energy and emergency services.

Protecting data infrastructure relies in part on protecting power and maintaining energy security. This has exposed a new reality: as data and energy become strategic dependencies, both are matters of national security.

As further proof of this, data centers and critical energy sites are known to be targeted in times of international conflict; such actions can exacerbate ongoing supply disruption and price volatility.

Against this backdrop, it’s clear that systems built on concentrated energy sources, such as an overreliance on one fuel type, can leave countries and operators exposed.

At the same time, it’s impossible to ignore the speed at which AI-driven energy demand is climbing. Data Centers are already consuming 6% of electricity in the UK and US, and that growth is not anticipated to slow.

However, because AI workloads are unevenly distributed and highly variable, data centers can have unpredictable fluctuations in power needs, surging during processing, dipping when they pause to share data or sync, and then repeating that cycle when they start the next batch. While daily averages might look consistent, demand throughout the day can fluctuate by hundreds of megawatts every few minutes.

In an environment of rising demand, unpredictable and dynamic load patterns, and geopolitical risk, simply adding more generation will not be enough. The real challenge is flexibility.

Flexibility will define the next era of energy

There is no single technology that can deliver the secure supply that AI will require. Resilience comes from combining dispatchable generation, renewable energy sources, storage, and smart energy management into a system flexible enough to respond to growing demand while guarding against volatility and disruption.

A data center’s ability to consistently access energy depends on the electricity system’s ability to adapt to changing conditions in real time, balancing supply and demand while maintaining a steady flow of power. Generation provides the power. Flexibility determines how effectively that power can be delivered and used, acting as a shock absorber against swings in supply and demand.

Historically, electricity systems have been designed around relatively predictable demand patterns.

Today, AI is introducing a new kind of demand. This dynamic is the reason flexibility will become one of the defining aims for the energy industry in years to come.

Flexibility as a business differentiator

Technology companies like Google and Meta are already diversifying to ensure energy security for their facilities. Google, for example, is purchasing nuclear power from Kairos Power’s small modular reactors to serve its AI facilities, while Meta has introduced a space-solar partnership with Overview Energy.

These investments demonstrate how energy resilience and flexibility are emerging as commercial differentiators for operators. Reliable access to power can be the difference between a data center’s success or failure. It allows organizations to manage risk, control costs, and scale more confidently. Flexibility can come from both the supply side (storage, dispatchable generation, and intermittent renewable generation) and the demand side (smart energy management, shifting workloads).

• Energy Storage: Captures electricity when supply is abundant to be used later on when demand is high. Solutions include battery energy storage systems and pumped storage hydro.
• Dispatchable Generation: Can ramp up quickly to generate electricity when needed. Examples include biomass and natural gas power plants.
• Smart Energy Management: Using advanced software and AI, virtual power plants and other solutions can optimise cooling systems, server utilisation, and power consumption in real-time based on grid conditions.
• Shifting Workloads: Moves non-urgent workloads to off-peak periods or physical locations to reduce stress on energy systems when demand is higher.

Together, these approaches can bolster energy systems and enable them to better respond to AI’s increasingly dynamic demand, while insulating against volatility and price swings.

Where We Stand Today

The good news is that there are signs of meaningful progress. In the UK, battery storage capacity surpassed nuclear generation for the first time, leaping from just under 10 MW to almost 7 GW in less than a decade. Elsewhere, earlier this year Emerald AI and the National Grid joined with other partners to complete a trial which proved that, through smart software, AI facilities can adjust electricity consumption in response to real-time demand signals without impacting critical operations.

Data centers need more of these solutions quickly because the cost of getting system flexibility wrong can be high, including rising costs, renewable curtailment, and reduced energy security. Meanwhile, as the geopolitical shocks that rocked energy markets have shown, reverting to and relying on the cheapest or fastest deployed single source also carries great risk.

AI’s energy demand is often framed as a race for more megawatts, but that’s an oversimplification. AI’s growth will depend just as much on the electricity system’s ability to adapt and continue delivering as it will on the megawatts.

To build a secure energy system capable of supporting the next generation of AI infrastructure, the UK must invest in flexible technologies that can meet the dynamic needs of future energy demand.

# # #

About the Author

Bio: Paul Sheffield is Chief Commercial Officer at Drax, where he leads the group’s biomass sales, trading and optimisation, and customer businesses. He joined Drax as Chief Operating Officer of Haven Power in 2017, bringing more than 20 years of experience in energy trading and retail, and was appointed to Drax’s Executive Committee in 2019.

The post The Next Big Challenge for Data Centers Isn’t More Power appeared first on Data Center POST.

TL;DR

• Selecting a data center UPS system requires evaluating load capacity, runtime, battery technology, serviceability, scalability, and long-term lifecycle costs rather than simply comparing initial purchase prices.
• Organizations should maintain a 20-30% spare capacity to support future growth, with a runtime of 5-15 minutes being sufficient to bridge the gap between utility failure and generator startup.
• Utilizing modular UPS configurations simplifies maintenance and reduces downtime risks, while focusing on the total cost of ownership (TCO) ensures optimal long-term reliability and efficiency.

# # #

Over the years, with the rapid advent of digital transformation across different industries, businesses’ data centers have been under pressure to deliver uninterrupted power, support growing workloads, and maintain operational efficiency. Reliable power protection remains a critical component of organizational continuity, whether supporting cloud services, enterprise applications, colocation environments, or AI-driven workloads.

The most important protection strategy is an Uninterruptible Power Supply (UPS) system, and choosing the right one is more than just comparing purchase prices. It is essential to evaluate UPS options for a data center based on load capacity, runtime, battery technology, serviceability, and lifecycle cost. A UPS must be evaluated based on its ability to support current and future load requirements, provide adequate runtime during outages, reduce maintenance needs, and deliver long-term value throughout its operational life.

When assessing UPS solutions, industry experts recommend a holistic approach and suggest organizations include the Uptime Institute, IEEE, and leading data center infrastructure providers.

Why Is UPS Selection Critical for Modern Data Centers?

For businesses, costly downtime, data loss and business interruptions can be highly affected by power disruptions. Against utility failures and power quality issues, UPS systems serve as the first line of defence by buffering and delivering clean, consistent power to protect the system from fluctuations. Apart from this, modern UPS solutions contribute towards:

• Availability and improved uptime
• Enhancement of power quality
• Better energy efficiency
• Increased operational stability
• Support for future growth of infrastructure

How Should Data Centers Evaluate Load Capacity?

The UPS system you choose should support the current load and future growth, and understanding load capacity is one of the most important factors to consider. Maintaining an approximately 20-30% spare capacity is often recommended to accommodate infrastructure expansion and avoid operating the UPS continuously at maximum capacity.

Here are some of the key considerations:

• Requirements of current IT power
• Growth of future workload
• Additions of planned equipment
• Target capacity utilisation
• Strategies of business expansion

How Much Runtime Does a Data Center Really Need?

During a power outage, the runtime refers to the amount of time a UPS can continue supplying power. Based on the business continuity strategy, the right runtime largely depends on it. While a UPS was never designed to provide power for hours, it bridges the gap between utility power loss and generator setup and a runtime of 5-15 minutes is considered sufficient.

While understanding the runtime requirements, it is important to consider the following:

• Availability of generator
• Plans of disaster recovery
• Requirements of critical applications
• Continuity of business
• Acceptable downtime

Which Battery Technology Is Best?

Reliability, maintenance requirements, space utilisation, and cost of ownership are directly affected by the battery technology decisions:

Valve-Regulated Lead Acid (VRLA) Batteries

Being a standard choice for data center UPS systems, VRLA batteries offer advantages like:

• Low acquisition costs
• Available within a broad market
• Maintenance-free operation
• Proven reliability
• Instantaneous response

Some limitations that should be considered:

• Shorter lifespan, typically 3-5 years
• High temperature sensitivity
• Larger physical footprint
• High operational costs

Quick Facts on Lithium-Ion Batteries

In modern data centers, lithium-ion batteries are becoming increasingly popular for good reason; here’s why:

• Extended lifespan, generally 10-15 years
• Space-saving footprint
• Quick recharge capabilities
• Less hands-on upkeep
• Operate safely in higher temperatures

Why Do Serviceability and Scalability Matter?

The ease of maintaining and expanding a UPS system directly influences its uptime and operational flexibility. Traditional systems can pose challenges, as maintenance activities require planned downtime or complex arrangements.

Modular systems provide a scalable approach to aligning infrastructure investments with business needs for organizations experiencing rapid growth or evolving workload demands.

Having modular UPS architectures offers multiple benefits, such as:

• Maintenance procedures are simple
• Fault isolation is improved
• Downtime risks are reduced
• Capacity expansion is easier

How Should Organizations Evaluate Lifecycle Cost?

The lifecycle cost analysis of organizations must include:

• Investment in initial equipment
• Expenses on battery replacement
• Performance of energy efficiency
• Maintenance costs preventions
• Requirements for cooling
• Agreements of service
• Risks involved in downtime
• Costs of equipment disposal

Final Thoughts

It is concluded that choosing a UPS system for a data center involves more than just upfront costs; it requires evaluating load capacity, runtime requirements, battery technology serviceability, scalability, and lifecycle costs. This ensures that the chosen solution aligns with business objectives and operational requirements.

Organizations that focus on total cost of ownership rather than acquisition cost alone are often better positioned to achieve long-term reliability, efficiency, and scalability. Industry leaders such as Schneider Electric continue to advance UPS innovation through high-efficiency power protection systems, modular architectures, and lithium-ion battery integration, helping data centers improve resilience while supporting future growth. Ultimately, the right UPS balances current needs with long-term operational and financial goals.

# # #

About the Author

Anuj Kudesia, MD-SEPSL and Director of Business Development at Schneider Electric, leads strategy, product launches, and revenue operations for the Secure Power Division across the Greater India Zone. With over 20 years at Schneider Electric spanning enterprise data center sales, channel management, and strategic business development including P&L ownership for BFSI, IT/ITES, and Government verticals, he brings deep expertise in UPS systems, critical power infrastructure, edge computing, and data center modernisation for Indian enterprises.

The post Choosing the Right UPS for Your Data Center: Load, Runtime, Battery Type, and Cost Compared appeared first on Data Center POST.

TL;DR

• The global hyperscale data center market is projected to skyrocket from USD 80.9 billion in 2025 to USD 624.2 billion by 2035, expanding at a compound annual growth rate (CAGR) of 22.2%.
• This explosive growth is heavily driven by AI-focused architectures, high-performance computing (HPC), and public cloud migrations, which require massive deployment of GPU-intensive server environments with higher rack power densities.
• Hardware accounted for 93.9% (USD 76 billion) of the market in 2025. While air-based cooling held a 52.8% share, direct-to-chip (DTC) liquid cooling is fast becoming the preferred thermal management method to handle intense AI workloads without requiring full infrastructure rebuilds.

 # # #

The global hyperscale data center market was valued at USD 80.9 billion in 2025 and is estimated to grow at a CAGR of 22.2% to reach USD 624.2 billion by 2035, according to the recent report by Global Market Insights Inc.

The market expansion is shaped by major changes in computing infrastructure, including the shift toward AI-focused architectures, high-performance computing environments, and increased dependence on cloud-based enterprise platforms. The growing need for advanced processing capacity for artificial intelligence applications is encouraging hyperscale operators to deploy GPU-intensive server environments with significantly higher rack power densities compared with traditional computing systems.

AI model development and large-scale data processing require advanced infrastructure capable of supporting high-performance workloads, creating substantial demand for hyperscale facilities. Enterprise digital transformation and the continued movement of business applications to public cloud platforms are further strengthening market growth. In addition, hyperscale companies are increasing investments in renewable energy procurement to support sustainability objectives and meet rising power requirements. The technology sector represented nearly 40% of corporate renewable power purchase agreements in 2025, highlighting the increasing role of hyperscale operators in clean energy adoption. Large-scale renewable energy agreements signed by major technology companies demonstrate a long-term approach toward securing reliable and sustainable energy sources for future data center expansion.

The hyperscale data center market from the hardware segment accounted for USD 76 billion, representing 93.9% share in 2025. This dominance is attributed to the significant capital requirements associated with hyperscale infrastructure development. Servers, storage systems, networking equipment, power management solutions, and cooling technologies collectively account for most of the investment required to establish and expand hyperscale facilities.

The air-based cooling systems segment held a 52.8% share, valued at USD 42.7 billion in 2025, owing to increasing thermal management requirements from AI-driven GPU infrastructure. Direct-to-chip (DTC) liquid cooling has emerged as the most adopted liquid cooling method, as it enables coolant circulation through cold plates attached directly to processors and memory components. This approach provides effective heat removal while allowing integration with existing data center designs without requiring complete infrastructure replacement.

North America hyperscale data center market held 42.8% share in 2025. The region continues to serve as the foundation of worldwide hyperscale investment, supported by significant infrastructure expansion from leading cloud and technology companies across major data center hubs. Continued capital spending by Amazon Web Services, Microsoft Azure, Google Cloud, Meta Platforms, and Oracle is contributing to the development of advanced hyperscale facilities across key markets, strengthening North America’s position in the global industry.

Key players operating in the global hyperscale data center market include Schneider Electric, Vertiv Group, Dell Technologies, Hewlett Packard Enterprise (HPE), Cisco Systems, Huawei Technologies, Super Micro Computer, Arista Networks, Lenovo, Eaton, Legrand, Hitachi, Gigabyte Technology, MiTAC Holdings, Cormant Technologies, Mortenson Construction, DPR Construction, Vast Data, LiquidStack, and GRC.

Companies operating in the hyperscale data center market are adopting strategies focused on infrastructure expansion, strategic partnerships, technological innovation, and sustainability initiatives to strengthen their market position. Industry participants are investing in advanced cooling solutions, high-density computing infrastructure, renewable energy integration, and next-generation networking technologies to address increasing AI and cloud workload requirements. Companies are also expanding their global data center footprints through new facility construction, acquisitions, and collaborations with energy providers and technology partners. Developing energy-efficient designs, improving operational reliability, and enhancing data center automation capabilities are key priorities for market participants.

The post Hyperscale Data Center Market to Surpass USD 624.2 Billion by 2035 appeared first on Data Center POST.

Data Center POST had the opportunity to connect with Shahar Belkin, Chief Evangelist at ZutaCore. Starting in 2018, he has led the development and product organizations, contributing to the advancement of next-generation two-phase cooling solutions for high-power silicon and data center infrastructure. Since 2024, he has served as the company’s Chief Evangelist, representing ZutaCore’s technology vision across the global AI and high-performance computing ecosystem.

What does ZutaCore do?

ZutaCore^® delivers waterless, two-phase direct-to-chip liquid cooling for data centres. The technology removes heat directly from CPUs and GPUs using a non-conductive dielectric fluid rather than water. Our flagship solution, ZutaCore HyperCool^®, helps AI, cloud and enterprise data centres cool higher-power processors, increase compute density and cut cooling energy use in half.

Proven at scale in demanding data centre environments, HyperCool eliminates water risk at the server level and gives operators a safer way to move beyond air cooling. Its cold plates, dielectric fluid, manifolds, cooling distribution units and control software are engineered together as one thermal architecture rather than sold as separate parts. In AI cooling, performance depends on how effectively the entire system works together, rather than on any single component.

What problems does ZutaCore solve in the market?

AI and high-performance workloads are pushing CPUs and GPUs to power levels that traditional air cooling can struggle to manage efficiently. We help data centre operators address this by removing heat directly at the chip, supporting higher-density compute without bringing water inside the server.

HyperCool enables this through a waterless, closed-loop, two-phase architecture that uses a non-conductive dielectric fluid to absorb and remove heat from processors. As rack densities continue to rise, this helps operators reduce water risk, improve cooling efficiency and scale AI infrastructure more safely.

What are ZutaCore’s core products or services?

Our flagship product is HyperCool, a waterless, two-phase direct-to-chip liquid cooling solution used across AI, HPC, cloud, financial services, enterprise and colocation environments. Every part of HyperCool is engineered as one integrated thermal system rather than optimised in isolation. It follows the same systems-engineering approach a Formula 1 team applies to overall lap time rather than any single part of the car.

HyperCool uses cold plates mounted directly on CPUs and GPUs. Inside each cold plate, our dielectric fluid absorbs heat from the processor, boils into vapour and carries that heat away from the chip. The vapour then condenses back into liquid in the heat rejection unit and recirculates through the closed-loop system.

We have also introduced OmniTherm™, a two-phase cold plate designed for advanced server configurations, including reversed orientation, inverted mounting, back-side power delivery and non-standard layouts. It gives data centre operators and system designers more thermal flexibility for high-density AI, HPC and cloud infrastructure.

What markets do you serve?

We serve the digital infrastructure market. Our focus is on AI data centres, high-performance computing, cloud, enterprise, colocation and edge environments. We also work with performance-sensitive sectors such as financial services and high frequency trading.

In these environments, cooling can affect latency consistency and hardware performance. It can also influence whether firms can deploy dense compute in strategic colocation facilities close to exchanges and liquidity venues.

What challenges does the global digital infrastructure industry face today?

Data centres are trying to support far more compute while using energy, space and water more carefully. AI workloads are pushing chip power and rack densities beyond what legacy air cooling was designed to handle. Facilities that once planned for tens of kilowatts per rack are now planning for much higher densities, and the industry is already discussing megawatt-class racks.

Power and cooling capacity are now also shaping growth, so operators need systems that use less energy, keep water away from IT equipment and reduce deployment risk. Yet much of the industry still discusses cooling in terms of individual parts, whether the cooling distribution unit, cold plate or coolant. The reality is that as power and density climb, cooling has become a systems-engineering challenge, where the design of each component affects the performance of the whole system.

How is ZutaCore adapting to these challenges?

We are scaling HyperCool from rack-level deployments to end-of-row and larger cooling systems capable of supporting higher-density AI infrastructure. By removing heat at the processor through a sealed, waterless two-phase loop, HyperCool helps operators reduce cooling energy and keep water away from IT equipment.

We are also validating multi-megawatt operation and designing systems around the practical requirements of live data centres. At our facility in Israel, we have built a 2 MW end-of-row emulation platform that replicates real-world thermal and facility conditions without using live IT equipment. This allows us to validate performance and stability at multi-megawatt scale ahead of customer deployment. OmniTherm extends the same approach to advanced server designs, including reversed orientation and back-side power delivery, giving operators more flexibility as AI hardware becomes denser.

What can we expect to see/hear from ZutaCore in the future?

We will use our recent $100 million Series C funding round to expand global commercialisation and advance R&D for next-generation AI and HPC chip cooling. The funding will also support work on in-package thermal management, megawatt-class deployments and integration with existing air and single-phase cooling systems.

We have more than 75 deployments worldwide today, and we are focused on scaling our platform as operators look to support higher compute density, reduce water risk and meet rising AI infrastructure demand. Our two-phase platform is designed to support next generation AI and HPC processors exceeding 4,000 watts. OmniTherm now supports full-power operation for the NVIDIA RTX PRO™ 6000 Blackwell Server Edition in a single-slot PCIe form factor. We will continue to advance the platform to meet the evolving requirements of AI-driven data centres.

What upcoming industry events will you be attending?

We will then be a Gold Sponsor at DataCloud USA in Austin, Texas on 2 and 3 September, where we will moderate the panel “Keeping it cool: evaluating air, liquid and hybrid approaches for high-density workloads”.

ZutaCore will also be a Sapphire Sponsor at the OCP Global Summit in San Jose, California from 12 to 15 October. On 15 October at 1:05pm, we will host the panel “Reducing Operational Complexity through Standard Two-Phase Liquid Cooling for AI Infrastructure” with Wiwynn, ASRockRack and Corinitis. We will also attend SuperComputing’26 in Chicago, Illinois from 15 to 20 November.

Do you have any recent news you would like us to highlight?

We recently announced our $100 million Series C funding round, with investment from Mitsubishi Electric, Carrier Ventures, Samsung Electronics through Samsung Ventures and additional investors. The funding will support global commercialisation, further R&D and continued work on cooling technologies for next-generation AI and HPC processors.

We also announced more than 75 deployments worldwide across the Americas, Europe and Asia, along with senior hires across finance, people, research and development and operations as we scale for further global growth. Those hires include Yaniv Reinhold as Chief Financial Officer, Sarah Warshavsky Oberman as Chief People Officer, Yoni Nir as Chief R&D Officer and Sharon Shafran as Chief Operating Officer, joining CEO Erez Freibach, President and CRO Brian Lillie, Chief Product and Technology Officer My Truong and Chief of Staff Susan Mor.

Is there anything else you would like our readers to know about ZutaCore and capabilities?

HyperCool is more than cold plates and fluid. It includes HyperCool Cloud, our software layer for real-time monitoring, control and optimisation across racks, rows and sites. The full system, with its hardware and software working together, is what customers are deploying.

ZutaCore was named an IDC Innovator in 2025, won Best Supercomputing Cooling Innovation at ISC 2025, and is a member of the Sustainable Digital Infrastructure Alliance and the STAC Alliance. Our $100 million Series C, backed by strategic investors including Mitsubishi Electric, Carrier Ventures and Samsung Electronics, reflects strong confidence in this approach as AI infrastructure moves to megawatt-class density.

Where can our readers learn more about ZutaCore?

Readers can visit the ZutaCore website and follow ZutaCore on LinkedIn to learn more about our technology, company news and upcoming events.

# # #

About Data Center POST

Data Center POST provides a comprehensive view of the digital infrastructure landscape, delivering industry insights into the global data center ecosystem. As the industry’s only peer-contributed and online publication, we offer relevant information from developers, managers, providers, investors, and trendsetters worldwide.

Data Center POST works hard to get the most current information and thought-provoking ideas most apt to add relevance to the success of the data center industry. Stay informed, visit www.datacenterpost.com.

If you are interested in contributing to Data Center POST, contact us at contributions@datacenterpost.com or submit your article here.

Want more digital infrastructure news? Stay in the know and subscribe to Data Center POST today!

The post Company Profile: How ZutaCore is Scaling Global AI Infrastructure appeared first on Data Center POST.

TL;DR

• In 2026, AI is shifting cloud ERP from a retrospective system of record that merely stores data into a system of prediction and autonomous action.
• Businesses achieve the most immediate, measurable returns by focusing on intelligent process automation (such as automated invoice matching), predictive financial forecasting to continuously update baseline models, and predictive inventory management to prevent supply chain stockouts.
• Successful deployment requires organizations to audit their data quality first, target two or three high-impact processes for initial pilots, establish clear boundaries for what AI can decide versus what requires human sign-off, and continuously check for model drift.

# # #

Most finance and operations teams don’t wake up asking for “AI in their ERP.” They wake up frustrated that month-end close still takes eight days, that inventory forecasts are wrong every quarter, and that someone still has to reconcile three systems before a board meeting manually. That’s the real starting point for this topic, not the buzzword.

Cloud ERP software has quietly become the operating backbone of most mid-size and enterprise businesses. What’s changed in 2026 isn’t the ERP itself – it’s what’s running underneath it. AI is moving from a bolt-on reporting feature to something closer to a coworker inside the system: reading documents, flagging anomalies, forecasting cash, and in some cases, taking action without waiting for a person to click “approve.”

This guide covers what that actually looks like: what’s genuinely new, where the real ROI shows up, and what to check before committing budget to an “AI-powered” ERP upgrade.

Key Takeaways

• AI-powered cloud ERP software combines the core system of record (finance, inventory, HR, supply chain) with machine learning, natural-language interfaces, and increasingly, autonomous AI agents.
• The highest-ROI use cases in 2026 are financial forecasting, predictive inventory management, and process automation, not chatbots.
• Agentic ERP workflows, AI that acts rather than just answers, are the biggest structural shift from the last ERP generation.
• The main adoption barrier isn’t the AI itself. It’s data quality, integration debt, and change management inside the business.
•  Judge any “AI-powered ERP” claim against three questions: what decision does it actually improve, what data does it need to work, and what happens when it’s wrong.

What Is Cloud ERP Software?

Cloud ERP software is a hosted platform that unifies core business functions, including finance, accounting, inventory, procurement, human resources, and supply chain, into a single system accessible over the internet rather than installed on local servers.

Instead of maintaining separate spreadsheets or on-premise systems for each department, a cloud ERP software gives every team a shared source of truth: one general ledger, one inventory count, one employee record. Vendors host the infrastructure, push updates automatically, and typically charge on a subscription basis.

Common examples include SAP S/4HANA Cloud, Oracle Fusion Cloud, Microsoft Dynamics 365, Oracle NetSuite, and Workday, all of which have added AI capabilities on top of this same core structure over the past two to three years.

How AI Is Changing Cloud ERP Software

AI changes cloud ERP in three layers: it makes existing processes faster through automation, it makes predictions more accurate through forecasting, and it makes the system usable by people who aren’t ERP specialists through natural-language interfaces.

Before AI, cloud ERP was primarily a system of record that stored data and let people query it. Reports were retrospective, so a business would find out inventory was short after the shelf was already empty. AI shifts ERP toward being a system of prediction and, increasingly, a system of action, flagging the shortage before it happens or reordering automatically within limits a human has set.

Analyst coverage over the past two to three years has consistently pointed the same direction: AI adoption inside ERP is accelerating faster than in most other enterprise software categories, with forecasting, reconciliation, and anomaly detection cited most often as the use cases delivering measurable time savings first.

Top AI-Powered Cloud ERP Platforms in 2026

Treat this table as a starting point, not a verdict. The right platform depends on your existing tech stack, industry, and which processes you actually want AI to touch first.

How to Evaluate an AI-Powered Cloud ERP Platform

Choosing between AI-powered ERP platforms comes down to a handful of criteria that matter more than any feature list, since most vendors now offer some version of automation, forecasting, and a conversational assistant.

• Data security and compliance: Confirm certifications relevant to your industry, such as SOC 2 or ISO 27001, and where the AI models process and store your data, especially if you operate under GDPR, HIPAA, or similar regulations.
•  Integration with your existing stack: Check what pre-built connectors exist for your CRM, e-commerce, and logistics tools, and how much custom development an integration will realistically require.
• AI governance controls: Look for granular settings on what an AI agent can do without approval, a clear audit trail of automated actions, and an easy way to override or roll back a decision.
• Total cost of ownership: Add up the base ERP license, AI module add-ons, implementation, and training, since the AI layer is rarely included in the headline subscription price.
•  Vendor roadmap and update cadence: Ask how often AI features change and whether you can control when updates roll out, especially for a business that can’t tolerate workflow disruption mid-quarter.
• Support for AI-specific issues: Confirm who you actually contact when an automated forecast or reconciliation looks wrong, and what the response-time commitment is.
• A sandbox test with your own data: Request a pilot using a real slice of your transaction data before signing, since AI accuracy claims in a demo rarely reflect how a model performs on your specific data quality.

In short: evaluate the platform on how it handles your actual data and processes, not on how convincing the AI demo looks.

7 Key Ways AI Is Transforming Cloud ERP Software in 2026

1. Intelligent Process Automation

Intelligent process automation means AI-driven bots handle repetitive, rules-based ERP tasks, such as invoice matching, purchase order approvals, and data entry, without a person doing it manually each time.

Unlike older robotic process automation tools that just followed rigid scripts, today’s automation layers use machine learning to handle exceptions. An invoice with a mismatched PO number gets flagged and routed correctly instead of stalling the whole workflow. That’s the practical difference: less babysitting, fewer stuck queues.

2. AI-Powered Financial Forecasting

AI-powered financial forecasting uses historical transaction data, seasonality, and external signals to project cash flow, revenue, and expenses more accurately than manual spreadsheet models.

Finance teams that used to rebuild forecasts every month by hand now get a continuously updated baseline that adjusts as new data comes in. The finance team still owns the assumptions and the final call. The AI just removes the grunt work of rebuilding the model from scratch each cycle.

3. Predictive Inventory and Supply Chain Management

Predictive inventory management uses AI to forecast demand at the SKU level and flag supply chain risks before they cause a stockout or overstock.

This is where ROI tends to show up fastest for product-based businesses: fewer emergency reorders, less capital tied up in dead stock, and earlier warning when a supplier is likely to miss a delivery window based on historical patterns.

4. Conversational ERP and Natural-Language Analytics

Conversational ERP lets non-technical users ask questions in plain language, such as “what were our top five underperforming SKUs last quarter,” and get an answer without writing a report or waiting on the analytics team.

This matters more for adoption than it sounds. A lot of ERP value has historically been locked behind people who know how to build the right report. Natural-language interfaces open that same data up to managers who never learned the reporting tool.

5. AI Agents and Agentic ERP Workflows

Agentic ERP workflows are AI systems that don’t just answer questions or flag issues. They complete multi-step tasks on their own, within boundaries a business sets, such as reconciling a discrepancy, generating a draft purchase order, or routing an approval to the right person.

This is the biggest structural change from the last wave of AI in ERP. Earlier tools mostly assisted a human who was still doing the work. Agentic workflows shift some of that work to the system itself, with a human reviewing outcomes rather than performing every step.

6. Real-Time Business Intelligence

Real-time business intelligence means dashboards and alerts update continuously as transactions happen, instead of refreshing on a nightly or weekly batch cycle.

For a leadership team, that’s the difference between finding out about a cash flow problem in next Monday’s meeting versus getting flagged the moment the pattern starts. AI adds the layer that decides what’s worth surfacing, instead of leaving someone to notice it in a static report.

7. AI-Driven Risk, Security and Compliance

AI-driven risk and compliance tools continuously monitor transactions for fraud patterns, policy violations, and regulatory exposure, rather than relying solely on periodic audits.

This includes anomaly detection on expense reports and payments, automated checks against changing tax and regulatory rules across jurisdictions, and audit trails that are easier to produce because the system has been tracking exceptions all along instead of reconstructing them after the fact.

Cloud ERP Software vs AI-Powered Cloud ERP

The core system, including the general ledger, inventory, and HR records, doesn’t change. What changes is how much of the analysis and routine decision-making the system can do before a person has to step in.

Benefits of AI-Powered Cloud ERP Software

• Faster close cycles: Automated reconciliation and anomaly flagging cut down the manual chase during month-end close.
• More accurate forecasting: Continuously updated models reduce the lag between something changing and the forecast reflecting it.
• Lower operational overhead: Fewer people spend time on data entry, matching, and routine exception handling.
• Earlier risk detection: Fraud, compliance, and supply chain risks get flagged as patterns emerge, not after a quarterly audit.
• Broader data access: Natural-language querying means more people in the business can self-serve answers instead of waiting on a report.

None of these benefits are guaranteed just by turning AI features on. They depend on clean underlying data and processes that are actually being followed, which is exactly where most implementations struggle.

Challenges of Integrating AI With Cloud ERP Software

• Data quality debt: AI forecasts and automations are only as good as the transaction data underneath them. Years of inconsistent entries don’t get fixed by turning on AI.
• Integration complexity: Most businesses run ERP alongside a patchwork of other tools, such as CRM, e-commerce, and logistics platforms, and AI features often need clean data flowing from all of them, not just the ERP itself.
• Change management: Teams that have run a process manually for years don’t automatically trust an AI recommendation, especially if the first few outputs are wrong.
• Governance and oversight: Agentic workflows that can take action need clear boundaries around what the AI is allowed to do without approval, and what always needs a human sign-off.
• Cost and licensing: AI modules are frequently priced as add-ons, and the total cost isn’t always obvious from the initial vendor pitch.

How to Prepare Your Business for AI-Powered Cloud ERP

Step 1: Audit your data quality first

Look at how consistent your chart of accounts, SKU records, and vendor data actually are. This is the single biggest predictor of whether AI features will work.

Step 2: Pick two or three processes, not everything at once

Month-end reconciliation, demand forecasting, or invoice matching are common starting points with measurable ROI.

Step 3: Define what “AI-assisted” versus “AI-automated” means for each process

Decide upfront what the system can do on its own and what still needs a human sign-off.

Step 4: Involve the people who do the work today

The team currently doing manual reconciliation or forecasting should help define what “correct” looks like. They’ll spot AI mistakes faster than anyone else.

Step 5: Run a pilot before a full rollout

Test the AI features on one business unit or process for a full reporting cycle before extending them company-wide.

Step 6: Set a review cadence

AI models drift as business conditions change, so build in a regular accuracy check rather than a one-time setup.

What Is the Future of Cloud ERP Software?

The direction is toward ERP systems that act more like a coordinated team of specialized agents than a single monolithic application: one agent handling procurement exceptions, another managing cash forecasting, another monitoring compliance, all working from the same underlying data and increasingly coordinating with each other.

Expect tighter integration between ERP and adjacent systems, such as CRM, HR, and supply chain planning, as vendors compete on how well their AI layer reasons across all of it, not just within one module. The businesses that get the most value won’t necessarily be the ones with the newest AI features. They’ll be the ones with clean data and clear rules for what the AI is allowed to decide on its own.

Conclusion

AI isn’t replacing cloud ERP. It’s changing what the system is capable of doing without a person driving every step. The businesses getting real value out of this shift aren’t the ones chasing every new AI feature. They’re the ones that fixed their data quality, picked a couple of high-impact processes, and set clear rules for what the AI can decide versus what still needs a human.

If you’re evaluating an AI-powered ERP upgrade, start with the question that actually matters: which specific decision, whether forecasting, reconciliation, inventory, or compliance, do you most need to make faster or more accurately? That answer should drive the platform and features you choose, not the other way around.

Considering an AI-powered ERP upgrade for your business? Talk to our team about which processes are worth automating first, and which aren’t.

Frequently Asked Questions

What is AI-powered cloud ERP software?

AI-powered cloud ERP software is a cloud-based enterprise resource planning system that uses machine learning, natural-language processing, and in some cases autonomous AI agents to automate tasks, generate forecasts, and support decisions across finance, inventory, HR, and supply chain, on top of the same core system of record as traditional ERP.

How much does an AI-powered ERP system cost?

Costs vary widely by vendor, company size, and which modules are deployed, and AI features are frequently priced as an add-on to the base ERP subscription. Get a detailed quote scoped to your specific processes rather than relying on published starting prices, which rarely reflect the full cost.

Is AI replacing ERP staff and finance teams?

No. AI is removing manual, repetitive work like data entry and reconciliation, but forecasting assumptions, exception judgment calls, and financial sign-off still require human oversight, especially as agentic workflows expand.

What’s the difference between AI-assisted and agentic ERP?

AI-assisted ERP surfaces insights and recommendations for a person to act on. Agentic ERP goes a step further and completes multi-step tasks on its own, within boundaries the business defines, with a human reviewing outcomes rather than performing every step.

Which businesses benefit most from AI-powered cloud ERP?

Businesses with high transaction volume, multiple locations or subsidiaries, and inventory-heavy operations tend to see the fastest ROI, since forecasting and reconciliation automation scale with volume.

What are the biggest risks of AI in ERP?

The main risks are acting on inaccurate data, since AI outputs inherit any underlying data quality problems, unclear governance over what AI agents can do without approval, and over-trusting automated outputs without a human review step.

Do I need to replace my existing ERP to get AI features?

Not necessarily. Most major ERP vendors are adding AI capabilities to their existing cloud platforms through upgrades or add-on modules, so a full replacement often isn’t required if you’re already on a modern cloud ERP.

How long does it take to implement AI features in an existing ERP?

Timelines depend on data readiness and process complexity, but a focused pilot on one or two processes typically takes a few months, while a company-wide rollout across all AI modules can take considerably longer.

The post How AI Is Transforming Cloud ERP Software in 2026 appeared first on Data Center POST.

TL;DR

• Leadership appointment: Dipan Patel joins Duos as Chief Operating Officer, reporting to CEO Doug Recker.
• Operational focus: Patel will lead deployment of Duos Edge AI data centers and GPU-as-a-Service platforms nationwide.
• Industry experience: He brings leadership experience across data centers, fiber networks, communications infrastructure, and technology strategy.

Duos Technologies Group, Inc. (“Duos” or the “Company”) (Nasdaq: DUOT) has appointed Dipan Patel as Chief Operating Officer, effective immediately. Patel joins the Company’s executive leadership team as Duos continues to scale its modular edge data center, colocation, and technology infrastructure solutions business.

As COO, Patel will have operational accountability for the deployment of Duos Edge AI data centers and GPU-as-a-Service platforms across the United States. Working with the executive leadership team, he will help strengthen operational performance and build the capabilities needed to support Duos’ growth as a next-generation AI infrastructure platform.

“Dipan brings a unique combination of deep infrastructure, technology, and operational leadership experience that aligns directly with our Company’s next phase of growth,” said Doug Recker, CEO of Duos. “His track record of building and scaling businesses across data centers and communications infrastructure will be instrumental as we expand our Edge AI platform and execute on our growth objectives.”

Prior to joining Duos, Patel served as a Digital Infrastructure Solutions Executive at Telstra InfraCo, where he led product strategy, solution innovation, and commercial monetization across a national portfolio of data centers, fiber networks, exchange facilities, and critical communications infrastructure. He previously served as Executive Vice President of Strategy, Technology and New Business Initiatives at SBA Communications Corporation, where he established the company’s corporate growth division and led investments in data centers, power solutions, and open-access fiber networks.

“AI is driving a once-in-a-generation infrastructure buildout, and Duos already has a proven modular platform and a team that can move fast,” said Dipan Patel. “I look forward to leveraging my background to help scale our operations, execute on our growth strategy, and expand the Duos footprint.”

Patel holds a PhD in Digital Communications and a BEng degree in Information Systems Engineering from Imperial College London. He is a Senior Member of the IEEE and holds 12 granted domestic and international patents.

The appointment reflects Duos’ continued investment in operational leadership as the Company expands its Edge AI-powered digital infrastructure ecosystem.

To learn more about Duos Technologies Group, visit duostechnologies.com.

The post Duos Technologies Group, Inc. Names Dipan Patel Chief Operating Officer appeared first on Data Center POST.

TL;DR

• While cloud computing has traditionally been defined by speed, scale, and flexibility, mature environments now present unpredictable costs, complex governance, and rising sovereignty concerns.
• Abstraction at scale has created a disconnect between organizations and the systems they must manage, with pricing hikes and licensing changes forcing businesses to re-evaluate control versus convenience.
• The future of cloud is defined not by how much an organization can scale, but by how effectively it can scale while retaining control, choice, and sovereignty.

# # #

By Juan Aguirre, CCO at Ilkari

Cloud computing has been defined by scale, speed, and flexibility for over a decade. As data volumes grew exponentially, organisations turned to the cloud to support growing workloads, enable rapid expansion, and to keep pace with increasingly dynamic industries.

But as cloud environments continue to mature, a new reality has emerged: flexibility alone is no longer enough. Today, many organisations face a very different data challenge. While traditional cloud systems still deliver in principle, how they work has ultimately changed how organisations operate and make decisions.

Costs are now more difficult to predict, governance is more complex to enforce, and sovereignty concerns are shaping infrastructure decisions in ways we’ve never seen before. What was once a decision based on convenience versus control is now a strategic business decision.

But the issue isn’t that the cloud has failed, it’s that abstraction at scale can create a disconnect from the systems organisations need to manage effectively.

Pricing hikes, changes in licensing, and growing dependence on third-party platforms are driving businesses to re-evaluate the balance between control and convenience

For steady, regulated, and performance-sensitive workloads, the answer is not more abstraction, but more selectivity. Organisations need cloud environments that are flexible where it matters, but bounded where control, predictability, and sovereignty are essential.

That requires designing infrastructure with cost control and governance built in from the start, rather than layered on later as an afterthought. It means recognising that not every workload needs unlimited elasticity. And it means creating cloud models that preserve choice over time, rather than quietly restricting it.

The next phase of cloud will not be defined by how much can be scaled, but by how effectively organisations can scale without losing control.

# # #

About the Author

​​Juan Aguirre is Ilkari’s Chief Commercial Officer, responsible for driving commercial strategy, partner ecosystems and customer success. With more than 25 years’ experience across Internet Protocol Television (IPTV), enterprise networking and SaaS, Juan has a strong track record of aligning technology offerings with evolving customer needs.

Before joining Ilkari, he served as EMEA Director for Partner Sales at CommScope, overseeing EMEA distribution and vertical markets such as hospitality and multi-dwelling unit (MDU). At Ilkari, Juan shapes the company’s go-to-market strategy, ensuring clients gain competitive advantage through tailored, forward-thinking solutions. His commercial leadership is central to Ilkari’s growth and its commitment to customer-centric innovation.

The post From Cost Drift to Sovereignty Demands: Why the Future of Cloud Is Selective appeared first on Data Center POST.

TL;DR

• Revenue growth: Second-quarter revenue increased nearly 30% year over year to $6.18 million, driven by the initial ramp in Technology Solutions revenue.
• Commercial progress: Duos signed 55 MW of hosting agreements with Axe Compute, secured a $111 million hyperscaler agreement in Columbus, Georgia, and now has 25 MW contracted for deployment in 2026.
• Outlook reaffirmed: The Company expects to deploy 25 MW this year and generate more than $50 million in 2026 revenue.

# # #

Duos Technologies Group, Inc. (“Duos” or the “Company”) (Nasdaq: DUOT), reported second-quarter 2026 results, reflecting continued progress in its transition to a focused provider of AI and edge data center infrastructure.

Revenue for the quarter ended June 30, 2026 increased nearly 30% to $6.18 million, compared with $4.77 million in the same period last year. The increase was driven primarily by the Company’s Technology Solutions business, which generated approximately $3.23 million in quarterly revenue through sourcing, integration, and supply chain services that support data center, AI, and enterprise computing deployments.

Duos also reported its first positive operating quarter as a data center infrastructure company, with approximately $0.05 million in operating income. Gross margin increased 94% year over year to $3.45 million, and adjusted EBITDA was positive at $0.5 million.

“Financially, we began to see the early stages of the substantial performance ramp we expect to build over the course of this year, highlighted by a 30% increase in revenue and a material improvement in profitability,” said Doug Recker, CEO of Duos. “We also secured over $100 million in growth capital through two major transactions.”

The Company continued to expand its commercial momentum during the quarter and in the weeks that followed. Duos signed five-year, 55 MW hosting agreements with Axe Compute valued at more than $500 million and secured $111 million in contracted revenue through a five-year agreement with an investment-grade hyperscaler for 10 MW of critical IT-load capacity at its Columbus, Georgia campus.

Duos now has 25 MW contracted, with all 25 MW planned for deployment in 2026. The Company also entered into an exclusive, non-binding term sheet with 0Lat LLC for a proposed structured lease across a 15-site, 225-cabinet edge data center portfolio in Texas and Georgia.

During the quarter, Duos secured more than $100 million in growth capital through a $55 million registered direct offering and $50.4 million in proceeds related to the sale of substantially all assets of New APR Energy, LLC. These proceeds support the Company’s planned Columbus campus expansion, related infrastructure investments, and contracted customer deployments.

Following the sale of its rail technology subsidiary, Duos is now fully focused on scaling its Edge Data Center and AI infrastructure businesses through Duos Edge AI and Duos Technology Solutions. The Company reaffirmed its 2026 expectation for revenue exceeding $50 million and 25 MW of deployed capacity by year end.

To learn more about Duos Technologies Group, visit duostechnologies.com.

The post Duos Technologies Group, Inc. Reports Q2 2026 Results appeared first on Data Center POST.

TL;DR

• The rapid expansion of AI computing has significantly increased energy consumption and altered the thermal and power profiles of server racks, forcing operators to rethink traditional facility designs.
• Designers are transitioning from traditional metals and legacy materials to high-performance thermoplastics, especially in environments where metal is prone to electrical interference, corrosion, or flammability risks.
• These advanced materials are best suited for critical architectural and equipment functions, including airflow containment systems, rack doors, plenum partitions, cable management, and power distribution.
• Thermoplastics provide a lightweight, highly available, and cost-effective alternative that features excellent impact and chemical resistance, strict flame-resistance compliance, and high dielectric insulation to prevent short circuits.

# # #

By Dan Mesler, Market Development Manager at PLASKOLITE

The rapid explosion of AI intelligence has altered the infrastructure landscape, escalating requirements for computing power and cloud storage reliability. This allows for global data consumption to drive unprecedented demand for hyperscale facility expansion. However, as advanced AI workloads multiply, operators face severe physical constraints – AI chips consume massive amounts of energy and drastically alter the thermal and power profiles of server racks.

To survive in environments that grow hotter with every complex query, facility designers must rethink their traditional reliance on legacy materials.

A substantial portion of infrastructure investment is shifting toward high-performance thermoplastics. Materials such as polycarbonate, PVC, and CPVC are actively replacing traditional perforated metal and legacy materials. Thermoplastics perform equally well or better than metals while offering significant structural advantages, particularly in facility zones where metal cannot be used due to electrical interference, severe corrosion risks, or strict flame-resistance mandates.

Within modern facility layouts, specialized materials like polycarbonate, acrylic, HDPE, and polypropylene are ideally suited for critical architectural and equipment applications. Key applications where these thermoplastics are best suited are:

• Airflow and containment systems
• Door and panel glazing
• Rack doors and side panels
• Plenum partitions and ceiling tiles
• Cable management and power distribution

The widespread adoption of these materials is fueled by their unique combination of physical properties and holding up well in extreme environments. High-performance plastics are inherently lightweight, cost-effective, and highly available compared to custom metal fabrications. More importantly, they provide excellent impact resistance, stand up well to chemical exposure and high temperatures produced in servers, have critical flame resistance that meets strict UL and FM flammability ratings, and have high dielectric insulation to prevent electrical short circuits.

As the artificial intelligence economy scales, every square foot of real estate is at an absolute premium. Stepping back from the chip itself, it is clear that facility design must adapt alongside server hardware. By utilizing these high-performance plastics, data center operators can successfully mitigate the extreme thermal, chemical, and electrical challenges of the next computing era.

# # #

About the Author

Dan Mesler is the Market Development Manager at PLASKOLITE, where he focuses on expanding the company’s presence in the data center market and identifying new opportunities for thermoplastic solutions within the industry. With approximately 15 years of experience in the plastics industry, he brings a strong combination of product knowledge, market expertise, and customer-focused strategy to his role.

The post How Thermoplastics Support Data Center Workloads appeared first on Data Center POST.

TL;DR

• The rising adoption of modular data center technologies is driving market growth by lowering civil engineering, design, and turnkey contracting expenses by approximately 30% compared to traditional data centers.
• Rising energy demands are pushing operators to adopt renewable energy sources (such as wind and solar) and implement advanced cooling techniques like liquid and AI-driven cooling to optimize efficiency.
• Propelled by technology giants investing in digital infrastructure and expanded cloud computing services, the United States dominated the sector in 2024 with USD 3.6 billion in revenue.

# # #

As per the report by Global Market Insights, Inc. “Data Centre Power Market ’’ was valued at USD 70 billion in 2025, and is projected to grow at a CAGR of 9.1% from 2026 to 2035.

The Global Data Center Power Market, The increasing adoption of modular data centers is a major factor driving this growth, as these facilities offer enhanced efficiency, scalability, and flexibility. They are becoming a preferred solution to meet the rising demand for data processing and storage. Modular power systems integrate generators, power distribution units, and uninterruptible power supplies in prefabricated designs that reduce costs and energy consumption, leading to lower operational expenses and a smaller carbon footprint. As digital infrastructure expands, the need for robust and energy-efficient power solutions continues to grow.

Leading technology firms and internet companies have already embraced modular data center technologies to optimize costs, reduce energy consumption, and accelerate deployment. Compared to traditional data centers, modular facilities lower civil engineering costs and cut design, management, and turnkey contracting expenses by approximately 30%. The push toward low-latency data processing, driven by advancements in edge AI, 5G networks, and cloud computing, is further fueling demand for high-performance power solutions in data centers.