The future of AI is landing in Taiwan. At NVIDIA GTC Taipei at COMPUTEX, the world’s developers, researchers and industry leaders are converging to dive into the latest breakthroughs shaping every industry, covering topics spanning AI factories and scaling infrastructure to agentic and physical AI and more. 

Hear from NVIDIA founder and CEO Jensen Huang live on stage at Taipei Music Center on Monday, June 1, 11 a.m. Taiwan time. Tune in early to catch the GTC Live at Taipei 2026 pregame show, featuring lively conversations with industry leaders about the latest innovations in AI and accelerated computing.

This is the place to find all the latest — stay tuned to the blog for live updates.

Thursday, May 21, 9 a.m. PT

NVIDIA Wins COMPUTEX 2026 Best Choice Awards for Innovations Spanning AI Factories, Robotics and Autonomous Vehicles 

NVIDIA Vera Rubin NVL72, NVIDIA Jetson Thor and NVIDIA Alpamayo were honored across four categories at Asia’s premier technology and computer trade exhibition.

At this year’s COMPUTEX Best Choice Awards (BCA), NVIDIA today received honors recognizing its innovation in AI computing, integrated circuits and autonomous vehicle (AV) development.

The NVIDIA Vera Rubin NVL72 rack-scale AI supercomputer won a Golden Award and the Sustainable Tech Special Award; the NVIDIA Jetson Thor platform for edge AI and robotics won a Golden Award; and the NVIDIA Alpamayo open platform for AV development won the Vehicle Technology and Smart Cockpit Category Award.

Entries were evaluated on their functionality, innovation and market potential, showcased at the premier computer and technology trade exhibition.

Jensen Huang, founder and CEO of NVIDIA, will deliver a keynote at COMPUTEX on Monday, June 1, at 11 a.m. Taiwan time.

NVIDIA Vera Rubin NVL72 Takes Home COMPUTEX Awards

Securing a Golden Award and the Sustainable Tech Special Award, Vera Rubin NVL72 connects 36 NVIDIA Vera CPUs and 72 NVIDIA Rubin GPUs — unified by the sixth-generation NVIDIA NVLink Switch for scale-up — with ConnectX-9 SuperNICs and Spectrum-X Ethernet Photonics co-packaged optics switches for scale-out and scale-across, as well as BlueField-4 DPUs to accelerate data processing across storage and security. 

Vera Rubin NVL72 delivers up to 10x higher inference performance per watt and 10x lower cost per token. When paired with NVIDIA Groq 3 LPX, Vera Rubin NVL72 delivers up to 35x higher throughput per watt for trillion-parameter models.

Designed for agentic AI, reasoning and long-context workloads, it enables AI factories to scale intelligence inside the rack and across the data center with secure, continuously available deployment.

The Vera Rubin NVL72 sets the bar for scalability, resiliency and sustainable AI infrastructure. Its cable-free, hose-free, fanless modular tray design reduces assembly time from two hours to five minutes per compute tray. 

The system’s power shelves deliver 6x more onboard energy storage for intelligent power smoothing, protecting both the rack and the broader power grid from steep load swings. In addition, its 100% liquid-cooled architecture operates at 45 degrees Celsius, meaning it drops seamlessly into existing liquid-cooled data centers and enables ambient-air, dry-cooler designs that redirect power from cooling overhead into token generation.

More BCA Wins for NVIDIA Technologies

NVIDIA Jetson Thor won a Golden Award as the most powerful edge AI compute platform built for physical AI and autonomous robots. Powered by the NVIDIA Blackwell GPU architecture, it delivers up to 2,070 FP4 teraflops of AI performance — 7.5x the compute and 3.5x the energy efficiency of the previous NVIDIA Jetson Orin generation — in a compact module configurable between 40 and 130 watts. 

Already in production across hundreds of applications, Jetson Thor is built to bring generative AI to smart robots, industrial systems, medical devices and autonomous machines while maximizing run-time performance and memory optimization.

Plus, NVIDIA Alpamayo won the Vehicle Technology and Smart Cockpit Category Award for pioneering open, reasoning-based autonomous vehicle development. Alpamayo is designed to help developers tackle rare, complex long-tail driving scenarios — such as interpreting an ambiguous hand signal from a pedestrian, determining the right-of-way when traffic lights and road markings contradict each other, and safely passing an emergency vehicle parked partially in the lane ahead — which fall outside typical training experience

The Alpamayo open platform includes Alpamayo 1.5 and Alpamayo 1, 10-billion-parameter chain-of-thought reasoning vision language action models for AV research; AlpaSim, an open source, end-to-end simulation framework for high-fidelity AV development; and NVIDIA Physical AI Open Datasets, which include more than 1,700 hours of driving data across geographies and conditions.

Learn more about NVIDIA’s latest innovations at NVIDIA GTC Taipei, running June 1-4 at COMPUTEX.

The mission begins now.

GeForce NOW is dialing up the action with a blockbuster mix of spy thrills, high-speed racing and member rewards — plus eight new games joining the cloud this week, all ready to stream instantly.

Leading the drop: the 007 First Light Ultimate Membership Bundle, which brings a brand-new way to jump into one of the year’s biggest releases and discover James Bond’s reimagined origin story.

Rev the engines — the Horizon Festival is calling. Forza Horizon 6 is now available on GeForce NOW — bringing the series’ signature open-world racing and festival energy directly to members, wherever they play. Stream a world built for speed, style and freedom instantly across devices, with no downloads required.

Step Into the Cloud of Espionage

A mission worth accepting arrives. Starting today through Wednesday, June 10, 007 First Light is included with the purchase of a 12-month GeForce NOW Ultimate membership. Lock it in before the game’s launch, and it’ll be ready to play the moment it goes live on Wednesday, May 27 — no preloads, no waiting, no mission briefings required.

This is a new, original take of Bond’s story — before the tux fit perfectly and the martinis got specific. 007 First Light drops players into a sharp, cinematic origin story packed with globe-trotting espionage, close calls and choices that shape how the world’s most famous agent takes form. Expect stealth, spectacle and just enough chaos to keep things interesting.

Stream it all with GeForce RTX 50 Series GPU power in the cloud, with up to 5K high dynamic range and cinematic-quality streaming for Ultimate members. It’s high-end PC gaming, minus the high-end PC.

Redeeming the mission is simple: After purchase, get the game by going to the “Available to Redeem” section in the account portal, signing in to a Steam account and completing the redemption.

The game is for members to keep — and ready to play across devices at launch with GeForce NOW.

Welcome to the Festival

Forza Horizon 6 is the ultimate automotive playground — a shared world where racing, exploration and creativity come together. Every road, trail and skyline is designed to be driven, discovered and mastered.

Take on a wide range of events, from high-speed street races to off-road expeditions, or drop into live activities happening across the world. Build a garage of iconic cars spanning decades of automotive history, each brought to life with incredible detail and customization. The Horizon Festival is more than competition — it’s a celebration of car culture and music. Seamless online features connect players across events and challenges, making every session feel alive and unpredictable.

Powered by the cloud, GeForce NOW delivers the festival in stunning detail. Members experience ultrasmooth performance and vivid environments without needing to wait for installs or upgrading hardware. With technologies like NVIDIA DLSS enhancing performance and visual fidelity, every reflection, speed blur and finish-line sprint hits at full intensity.

Community Spotlight

Community-created content takes the spotlight this week, as GeForce NOW ambassador Cloud Gaming Battle sits down with Andrew Fear, GeForce NOW product manager director at NVIDIA, to dive into what makes the service tick. The conversation covers how GeForce NOW brings high-performance PC gaming from the cloud to a wide range of devices, giving a closer look at the tech, thinking and passion behind the platform. It’s a conversation that the GFN Reddit community is calling “genuinely inspirational.”

In addition, members can look for the following:

• Forza Horizon 6 (New release on Steam and Xbox, available on Game Pass, May 19)
• Deep Rock Galactic: Rogue Core (New release on Steam, May 20)
• Luna Abyss (New release on Steam and Xbox, available on Game Pass, May 21)
• Warhammer 40,000: Mechanicus II (New release on Steam, May 21)
• ZERO PARADES: For Dead Spies (New release on Steam, May 21)
• Splitgate Arena Reloaded (Xbox, Available on Game Pass)
• Sunderfolk (Epic Games Store)
• TerraTech Legion (Steam and Xbox, available on Game Pass)

What are you planning to play this weekend? Let us know on X or in the comments below.

Rev your engines — the Horizon Festival is calling. @ForzaHorizon 6 is now streaming on GeForce NOW – stream a world built for speed in the cloud today. pic.twitter.com/hJuKFT2GZp

— NVIDIA GeForce NOW (@NVIDIAGFN) May 19, 2026

At this year’s Google I/O conference, NVIDIA and Google Cloud are accelerating the work of more than 100,000 developers in the companies’ joint developer community, which provides curated learning paths, hands-on labs and events that help them build using the full-stack NVIDIA AI platform on Google Cloud. 

Launched at Google I/O last year, the community brings together developers, data scientists and machine learning engineers who want to sharpen their AI skills on the latest NVIDIA and Google Cloud technologies. 

New additions for the community are rolling out this year, including a learning path for using the JAX library on NVIDIA GPUs, a new NVIDIA Dynamo codelab focused on inference optimizations, as well as monthly developer livestreams. 

Over the last year, the community has become a go‑to hub for AI builders using NVIDIA‑accelerated tools for data science and machine learning. The result has been production‑ready retrieval-augmented generation applications on Google Kubernetes Engine (GKE) and instrumenting observability for agent workloads. 

These AI builders are also experimenting with new large language model research and prototyping hybrid on‑premises and cloud inference for real‑world use cases like sports analytics and enterprise data pipelines. 

Building With Google DeepMind’s Gemma, NVIDIA Nemotron and Open Frameworks

NVIDIA and Google Cloud are equipping developers with learning resources and hands-on labs that combine NVIDIA libraries, open models and tools with Google Cloud’s AI platform — so they can build optimized, production‑ready AI applications faster.

For example, developers can accelerate data science and analytics with the NVIDIA cuDF library in Google Colab Enterprise or Dataproc, or deploy multi-agent applications by combining Google DeepMind’s Gemma 4 models, NVIDIA Nemotron open models and Google Agent Development Kit with Google Cloud G4 VMs powered by NVIDIA RTX PRO 6000 Blackwell GPUs in Google Cloud Run or with spot instances. 

NVIDIA and Google Cloud work closely across open frameworks like JAX so developers can build, scale and productize JAX workloads on NVIDIA AI infrastructure on Google Cloud — from single‑GPU experiments to multi‑rack deployments — while getting strong performance and a consistent experience. 

This work extends to Google Cloud AI Hypercomputer, where the MaxText framework uses these JAX optimizations to train large models efficiently on NVIDIA GPUs.

Building on the same foundation, NVIDIA Dynamo on GKE helps developers optimize large-scale inference — including mixture-of-experts models — so they can serve AI applications more efficiently with NVIDIA accelerated infrastructure on Google Cloud.

To help developers get hands-on with these capabilities, a new learning path on running and scaling JAX on NVIDIA GPUs and a new NVIDIA Dynamo on GKE inference codelab will become available next month for members in the Google Cloud and NVIDIA developer community.

Advancing Responsible AI With Google DeepMind’s SynthID and NVIDIA Cosmos

AI agents are increasingly built from a system of AI models — combining proprietary and open source models that reason, plan and act on users’ behalf. 

Amid this shift, trust and transparency are foundational, so developers and organizations can understand how these systems work and what they generate.

NVIDIA was the first industry partner to collaborate with Google DeepMind on SynthID, an AI watermarking technology that embeds robust digital watermarks directly into AI‑generated content, which helps preserve the integrity of outputs from NVIDIA Cosmos world foundation models available on build.nvidia.com.

Cosmos models provide rich 3D perception and simulation capabilities for robots, autonomous machines and other physical AI systems, while SynthID brings content transparency to the imagery and video they rely on. 

Together, they help preserve the integrity of AI‑generated content so developers can build and deploy agentic applications more responsibly across cloud, edge and real‑world environments.

Building on a Full-Stack NVIDIA and Google Cloud Platform

This year, Google I/O is putting the spotlight on new agentic experiences and tools for developers — and NVIDIA and Google Cloud are focused on ensuring builders have the infrastructure, software and learning resources they need to make the most of them. 

For developers in the community building on NVIDIA and Google Cloud, the skills and tools they learn can scale, effortlessly taking projects from prototype to enterprise‑grade workloads. 

At Google Cloud Next, Google Cloud and NVIDIA expanded their full‑stack platform to help developers train, deploy and operationalize agents on Google Cloud. This collaboration includes work on NVIDIA Vera Rubin-powered A5X instances, Google DeepMind Gemini models and more, and is being harnessed by leading AI labs and enterprises including OpenAI, Thinking Machine Labs, Schrodinger, Salesforce, Snap and Crowdstrike. Learn more in this blog.

Join the NVIDIA and Google Cloud developer community to connect with other builders and stay up to date on new tools, developer events and programs.

Agentic AI inference at one-tenth the cost per token with NVIDIA Vera Rubin NVL72. Agent sandboxes run 50% faster on NVIDIA Vera than traditional CPUs — while enterprise data queries are up to 3x faster with the Vera CPU. And 5,000 enterprises like Lilly, Samsung and Honeywell are running AI workloads on Dell AI Factories with NVIDIA, turning ambition into production at scale.

That’s the picture Michael Dell painted Monday morning at Dell Technologies World. Dell sized the stakes: Worldwide AI infrastructure spending could reach $3-4 trillion by 2030, with token consumption projected to grow 3,400% in the same window.

“There is a massive AI investment boom thats already underway, and a productivity boom is beginning, and in some companies, including ours,” Dell said. “The rate of change has gone parabolic, and it’s not slowing down.”

Then, the Dell chairman and CEO welcomed NVIDIA founder and CEO Jensen Huang to the keynote stage — with a look at the NVIDIA portfolio behind him, from a deskside Dell Pro Max with GB10 workstation to a Dell PowerRack with NVIDIA Vera Rubin NVL72.

“We’ve now arrived at the era of useful AI, which is the reason why demand is going parabolic, utterly parabolic,” Huang said. “What took months now takes weeks. What took weeks now takes days. And what takes days now takes hours. It’s a big deal in productivity, but a gigantic leap in computation requirements.”

The message: Enterprise AI has moved past pilots into agentic AI and inference deployments at scale. The platform for what’s next is the Dell AI Factory with NVIDIA — running frontier models and autonomous agents securely behind the enterprise perimeter.

A New AI Factory for the Agentic Era

The accelerated computing news leads the refresh: The Dell PowerEdge XE9812, built on NVIDIA Vera Rubin NVL72, delivers up to 10x lower cost-per-token than NVIDIA Blackwell for massive-scale agentic AI inferencing.

It’s joined by PowerEdge XE9880L, XE9885L and XE9882L servers — the first Dell systems built on NVIDIA HGX Rubin NVL8, supporting up to 144 GPUs per rack with 100% direct liquid-cooled compute nodes and up to 10x the performance of HGX B200.

In addition, networking gets the new Dell PowerSwitch portfolio with NVIDIA Quantum-X800 InfiniBand, featuring liquid-cooled, co-packaged optics and NVIDIA Spectrum-6 Ethernet.

Dell is also introducing Dell PowerRack, a fully integrated system – compute, networking and storage engineered as one – with thermal design, power management and software optimization built to work together from the ground up. The result is accelerated AI and high-performance computing workloads at enterprise scale, without the integration overhead of component assembly. 

On the CPU side, Dell PowerEdge M9822 and R9822 servers bring NVIDIA Vera CPUs to the enterprise AI factory. Purpose-built for agentic AI, Vera runs data pipelines, analytics, sandboxed tools and code workloads where each step waits on the last. 

With 1.2 TB/s memory bandwidth and predictable performance under load, Vera completes agentic workloads 50% faster than x86 processors, helping PowerEdge systems increase AI factory output with faster agent responses and shorter feedback loops.

“Vera CPU has the highest single-threaded performance of any CPU in the world,” Huang said. “It has three times the memory bandwidth — as a result, Starburst, DuckDB, all these databases run incredibly fast, because the agents are pounding on the databases, so the CPU had better be super fast.”

Starburst, a new data engine in the Dell AI Data Platform with NVIDIA, delivers 3x faster query throughput on NVIDIA Vera CPU for large-scale SQL analytics. 

Enterprise data provides the fuel for the AI factory. Dell’s update to its AI Data Platform with NVIDIA centers on accelerated data engines built on NVIDIA CUDA-X libraries — including cuDF for structured data and cuVS for unstructured data. 

Multiple customers for Dell AI Factory with NVIDIA were featured in the keynote. 

Diogo Rau, executive vice president and chief information and digital officer of Lilly, joined early in the keynote — discussing Lilly’s AI-driven advancements and innovation in life sciences, powered by AI infrastructure deployed at scale with Dell and NVIDIA.

He described technology as key to delivering cutting-edge science, at scale. “I think we’re on the verge of maybe being able to end disease as we know it,” Rau said. “Something like that was completely unimaginable 20 years ago, but today we can imagine it.”

A video from Samsung followed — highlighting use cases for R&D chip design and manufacturing running on Dell AI Factory with NVIDIA. 

Honeywell chief technology officer Suresh Venkatarayalu joined Michael Dell to walk through the company’s move from public cloud to on-premises AI — using the Dell AI Factory and Dell AI Data Platform with NVIDIA for industrial AI use cases, digital twins and automation from data center to the edge.

“For me, partnering with Dell and NVIDIA is not just about getting infrastructure,” Venkatarayalu said. It’s the full AI stack, he explained: scalable, secured and trusted by customers.

And in financial services, Hudson River Trading, the algorithmic trading firm, is expanding its Dell deployment to power AI-driven research — running Dell PowerEdge XE9685L servers with NVIDIA accelerated computing and NVIDIA Spectrum-X Ethernet to scale with the firm’s data, models and ambition.

Agents and Models on Premises — Securely

Dell’s own AI adoption survey, cited from the keynote stage, found that 67% of AI workloads now run outside the cloud — on premises, on device, at the edge or in colocation — and that 88% of respondents are running at least one AI workload on premises. 

The on-premises AI announcements answered a question Dell put to the room directly: “How do you deploy the world’s best AI models where you need them, with security and governance built in?” 

The answer rests on NVIDIA Confidential Computing — delivered together with Fortanix, Google, Red Hat and other partners — as the foundation for securely deploying frontier models inside the enterprise without exposing model IP or data. 

This enables enterprises to protect AI models and sensitive data in use while capturing the token efficiency, performance and cost advantages of on-premises AI infrastructure.

Frontier Proprietary Models, Protected by Confidential Computing

Google Distributed Cloud (GDC) with Gemini 3.0 is now available in preview on Dell PowerEdge XE9780 servers, accelerated by NVIDIA Blackwell and secured by NVIDIA Confidential Computing — giving enterprises a private confidential computing environment for advanced AI.

SpaceXAI will also bring the latest SpaceXAI models on premises to the Dell AI Factory, with NVIDIA Confidential Computing keeping model weights and enterprise data protected end to end.

Open Frontier Models, Running Natively on the Dell AI Factory

NVIDIA Nemotron models — frontier-class open intelligence — run on Dell AI Factory infrastructure for enterprises that want open weight models tuned to their own domains and data. 

Reflection’s open source frontier AI models are also coming on premises, purpose-built for regulated industries, governments and sovereign entities. 

Additional open models — MiniMax-M2.7, DeepSeek Pro, DeepSeek-V4, GLM 5.1 and Kimi K2.6 with NVIDIA NVFP4 optimization — are available on the Dell Enterprise Hub on Hugging Face, joining Gemma 4, NVIDIA Nemotron Super 3, Mistral Small 4 and Arcee Trinity-Large-Thinking.

In this new agentic era, enterprises also need agents to work securely across the hybrid and on-premises environments where their data, systems and workflows already live.

OpenAI Codex will connect with the Dell AI Data Platform, to help customers bring Codex closer to the internal context that makes agents useful: codebases, documentation, business systems, operational knowledge and team workflows. Dell and OpenAI will also explore how Codex can connect with the Dell AI Factory.

Ecosystem of Software Partners

Dell announced several new software partnerships for common enterprise AI use cases ranging from agentic AI and code assistants to computer vision.

This includes Palantir’s sovereign AI OS reference architecture with NVIDIA, announced in March, which now runs on Dell infrastructure — for on-premises deployment of Palantir Ontology and AIP, integrated with the NVIDIA Sovereign AI OS Reference Architecture.

In addition, ServiceNow customers will be able to leverage the Dell AI Factory to bring together infrastructure and enterprise workflow automation, enabling organizations to discover, govern and operationalize AI focused on business outcomes.

Dell also announced new solutions with a wide range of AI leaders and software innovators —including Fogsphere, Ipsotek, Mistral AI, Poolside and Uneeq — as well as security partnerships with CrowdStrike and Fortanix.

Agents From Deskside to Data Center

The most personal news: Dell Deskside Agentic AI with the NVIDIA NemoClaw stack, NVIDIA OpenShell runtime and NVIDIA Nemotron open models run on Dell Pro Max with GB10 and GB300 powered by the NVIDIA Grace Blackwell architecture, as well as Dell Pro Precision systems powered by NVIDIA RTX PRO Blackwell workstation platforms.

The customization layer: NVIDIA Nemotron, NVIDIA Agent Toolkit and NVIDIA NeMoClaw — the agent orchestration harness Huang described on stage as the connective layer between local models and enterprise data — provide the foundation for building enterprise autonomous agents, enabling organizations to customize models, orchestrate agent workflows and securely connect agents to enterprise data and tools.

The security layer: NVIDIA OpenShell — an open source runtime for development and deployment of autonomous agents with security and privacy controls — enables corporate policy enforcement at the infrastructure layer and is integrated with leading enterprise software platforms.

NVIDIA OpenShell is now supported across the entire Dell AI Factory with NVIDIA, giving developers a secure runtime to build, deploy and govern AI agents from workstations to servers. 

Dell also highlighted support for the NVIDIA AI-Q Blueprint, giving enterprises a reference example to deploy multi-agent workflows for deep research — accelerating the path from development to pilot to production.

Day Two

On day two of Dell Technologies World, Dell’s Chief Operations Officer Jeff Clarke and Infrastructure Solutions Group President Arthur Lewis will go deeper on NVIDIA Vera CPU, Vera Rubin, Confidential Computing and Nemotron — with a live demo of Dell Deskside Agentic AI.

The themes Huang and Dell set Monday — safe, long-running agents, full-stack factories, secure on-premises deployment — set up the broader announcements NVIDIA will bring to GTC Taipei at COMPUTEX, running June 1-4. 

Editor’s note: The Gaijin single sign-on feature is now up and running.

Dive masks on — Subnautica 2 is making a splash on GeForce NOW day-and-date with launch, so members can plunge into the title’s brand-new alien ocean from almost any device.

It leads 11 new games joining the cloud this week.

A limited-time HITMAN World of Assassination reward event brings signature tools of the trade — equal parts precision and unpredictability. At the same time Le Chiffre from CASINO ROYALE, played by the legendary actor Mads Mikkelsen, returns to HITMAN World of Assassination. Play for free through May 25 and earn a unique reward for 007 First Light.

Meanwhile, engines are already revving — the highly anticipated Forza Horizon 6 races onto the scene with early access, starting today for players who preordered the Premium Edition on Steam or purchased the Premium Upgrade Bundle for PC Game Pass. GeForce NOW members can jump in once early access becomes available on the service — no preinstalls or downloads needed. 

It all highlights what GeForce NOW does best: turning PC gaming collections into play-anywhere experiences across devices, with no install, updates or the latest hardware required.                                                                                                                                                                                                                      

Making Waves

Jump into a shared-world ocean frontier in Subnautica 2 — where survival, exploration and creativity rise to the surface together. Every reef, trench and stormy horizon is crafted to be swum through, studied and slowly tamed as players chart their own course beneath alien waves.

Descend into a living seascape of dynamic ecosystems, mysterious ruins and creatures that range from curious to colossal. Take on story-driven missions, dive into resource-rich biomes and construct bases above and below the surface to turn hostile waters into a home. Whether playing solo or with a crew, the ocean stays vast and unpredictable.

On GeForce NOW, every current, glow and shadow streams with crisp detail and fluid performance straight from the cloud. Members can plunge into the depths in moments, skipping downloads and hardware upgrades. 

Keep It Classy

Suit up, agent — things are getting a little explosive.

A new HITMAN World of Assassination reward event lands on GeForce NOW, giving members a stylish (and slightly unhinged) way to make an impression.

Each item leans into HITMAN’s signature tone — equal parts precision, chaos and dark humor. Here’s what’s waiting:

• Free users: The Purple Streak Explosive Duck — a remote explosive disguised as an innocent rubber toy.
• Performance members: The Purple Streak Explosive Duck and the Bomb Dynamite, a classic TNT bundle built for loud, messy exits.
• Ultimate members: Full Ultimate Bundle — adds the Purple Streak Fiber Wire (a sleek, silent takedown tool) and the Purple Streak Swimwear Suit (a bold outfit that trades stealth for style) to the Purple Streak Explosive Duck and the Bomb Dynamite.

Starting today, all members get access to the rewards, which will remain available through Sunday, June 14, first come, first served. To claim, log in to a GeForce NOW account, head to the rewards section in the account portal and redeem. 

One mission, a handful of questionable tools and a clear message: Stealth is optional, style is not.

Splashy New Games

In addition, members can look for the following:

• Battlestar Galactica: Scattered Hopes (New release on Steam, May 11)
• Outbound (New release on Steam, May 11)
• Directive 8020 (New release on Steam, May 12)
• Subnautica 2 (New release on Steam, Epic Games Store and Xbox, available on Game Pass, May 14)
• Blades of Fire (New release on Steam, May 14)
• Forza Horizon 6 (Early access on Steam and Xbox, available on Game Pass, May 14)
• Disco Elysium (Xbox, available on Game Pass)
• Oddsparks: An Automation Adventure (Epic Games Store)
• Planet Coaster 2 (Epic Games Store)
• Sudden Strike 5 (Steam)
• Super Battle Golf (Steam)

The Gaijin single sign-on feature is coming soon and is currently in development with Gaijin. Updates will be shared in a future GeForce NOW Thursday once the feature is live.

What are you planning to play this weekend? Let us know on X or in the comments below.

Which gaming genres do you play the most?

— NVIDIA GeForce NOW (@NVIDIAGFN) May 13, 2026

Reinforcement-learning agents — AI systems that learn by trial and error — can convert computation into new knowledge.

That’s the focus of a new engineering-level collaboration between NVIDIA and Ineffable Intelligence, the London-based AI lab founded by AlphaGo architect David Silver in the wake of Ineffable’s emergence from stealth last week.

“The next frontier of AI is superlearners — systems that learn continuously from experience,” said Jensen Huang, founder and CEO of NVIDIA. “We are thrilled to partner with Ineffable Intelligence to codesign the infrastructure for large-scale reinforcement learning as they push the frontier of AI and pioneer a new generation of intelligent systems.”

Silver is one of the pioneers of reinforcement learning, an approach that has transformed AI research. He’s focused on further developing this approach into a new paradigm.

“Researchers have largely solved the easier problem of AI: how to build systems that know all the things humans already know,” Silver said. “But now we need to solve the harder problem of AI: how to build systems that discover new knowledge for themselves. That requires a very different approach — systems that learn from experience.”

That kind of learning needs a powerful and highly optimized pipeline to support it. Unlike pretraining, where a fixed dataset of human data flows through the system, reinforcement learning workloads generate their data on the fly. 

The system has to act, observe, score and update continuously in tight loops, which puts pressure on interconnect, memory bandwidth and serving in ways that pretraining doesn’t. Furthermore, the system will train on rich forms of experience that are quite distinct from human language and other human data, and may require novel model architectures and training algorithms. 

That’s where NVIDIA and Ineffable are focusing their technical work: building a pipeline that can feed reinforcement learning systems at scale. Engineers from both companies have teamed up to explore the best way to create this training pipeline. 

This work is starting on NVIDIA Grace Blackwell, and will be among the first to explore the upcoming NVIDIA Vera Rubin platform. The goal is to understand the next generation of hardware and software that will be required as the AI world shifts beyond human data toward models that learn through simulation and experience. 

Getting this infrastructure right will unlock an unprecedented scale of reinforcement learning in highly complex and rich environments, allowing agents to discover breakthroughs across all fields of knowledge. 

Agentic AI is changing the way users get work done. Following the success of OpenClaw, the community is embracing new open source agentic frameworks. The latest is Hermes Agent, which crossed 140,000 GitHub stars in under three months and, as of last week, is the most used agent in the world according to OpenRouter.

Developed by Nous Research, Hermes is designed for reliability and self-improvement — two qualities that have historically been hard to achieve with agents. It’s provider- and model-agnostic by design, and optimized for always-on local use, making NVIDIA RTX PCs, NVIDIA RTX PRO workstations and NVIDIA DGX Spark the ideal hardware to run it at full speed, around the clock. 

Qwen 3.6, a new series of high-performance, open weight large language models (LLMs) from Alibaba, are ideal for running local agents like Hermes. The Qwen 3.6 27B and 35B parameter models are outperforming their previous-generation 120B and 400B parameter model counterparts and run on NVIDIA RTX and DGX Spark for accelerated agentic AI.

Hermes: Local AI Agent Capabilities Accelerated

Like other popular agents, Hermes integrates with messaging apps, can access local files and applications, and runs 24/7. But four standout capabilities set it apart:

• Self-Evolving Skills: Hermes writes and refines its own skills. Every time the agent encounters a complex task or receives feedback, it saves its learnings as a skill so it can adapt and improve over time.
• Contained Sub-Agents: Hermes treats sub-agents as short-lived, isolated workers dedicated to a sub-task — with a focused context and set of tools. This keeps task organization tidy, minimizes confusion for the agent and allows Hermes to run with smaller context windows, which is ideal for local models.
• Reliability by design: Nous Research curates and stress-tests every skill, tool and plug-in that ships with Hermes. The result: Hermes just works — even with 30 billion-parameter-class local models — without the constant debugging that most other agent frameworks require.
• Same model, better results: Developer comparisons using identical models across frameworks consistently show stronger results in Hermes. The difference is the framework: Hermes is an active orchestration layer, not a thin wrapper, enabling persistent, on-device agents instead of task-by-task execution.

Both the Hermes agent and the underlying LLM are built to run locally — which means the quality of hardware directly determines the quality of a user’s experience. NVIDIA RTX GPUs are purpose-built for this kind of workload.

Qwen 3.6: Data Center-Level Intelligence, Locally

The latest Qwen 3.6 models build on the acclaimed Qwen 3.5 series to deliver another leap forward for local AI agents. The new Qwen 3.6 35B model runs on roughly 20GB of memory while surpassing 120 billion-parameter models, which require 70GB+ of memory. 

In addition, Qwen 3.6 27B is a new, dense model with more active parameters — matching the accuracy of 400 billion-parameter models like Qwen 3.5 397B while being one-sixteenth the size. Running on high-end RTX GPUs provides the model the computing power it needs for a speedy experience.

These models are ideal for local agents like Hermes, and NVIDIA GPUs and DGX Spark are the fastest way to run them. NVIDIA Tensor Cores accelerate AI inference to deliver higher throughput and lower latency — so Hermes can work through a multistep task or refine one of its own skills in seconds rather than minutes.

DGX Spark: The Always-On Agentic Computer

Agents like Hermes are built to run continuously — responding to requests, planning multistep tasks, executing autonomously and self-improving. NVIDIA DGX Spark is the ideal companion — a compact, efficient standalone machine built for sustained, all-day agentic workflows. 

With 128GB of unified memory and 1 petaflop of AI performance, NVIDIA DGX Spark can run 120 billion-parameter mixture-of-experts models all day. And the new Qwen 3.6 35B model delivers equivalent intelligence in a leaner footprint — running faster and giving users the capacity to run concurrent workloads.

To maximize performance and ease of use, read the Hermes DGX Spark playbook. Plus, register for upcoming hands-on sessions in NVIDIA’s “Build It Yourself” agentic AI series to learn how to build autonomous AI agents with NemoClaw and OpenShell.

NVIDIA DGX Spark is available to order from NVIDIA’s manufacturing partners — visit the marketplace.

Getting Started With Hermes on NVIDIA Hardware

Running Hermes locally on NVIDIA hardware is straightforward. 

Visit the Hermes GitHub repository to get started, and pair it with a preferred local model and runtime. Run Hermes alongside Qwen 3.6 via llama.cpp, LM Studio or Ollama. Hermes Agent ships with LM Studio and Ollama support out of the box for the simplest path to a local agent.

Whether for a local AI enthusiast exploring the frontier of personal agents or a developer building local tooling for their workflows, Hermes on NVIDIA hardware offers a uniquely capable and reliable foundation.

Stay tuned for more updates from RTX AI Garage on the latest open models and agents optimized for NVIDIA RTX hardware.

#ICYMI: The Latest From RTX AI Garage

NVIDIA RTX PRO GPUs deliver up to 3x faster token generation running Qwen 3.6 models with llama.cpp. Get the real-time responsiveness needed for local AI, where agents can tackle multistep tasks and refine their skills to keep workflows seamless. 

Google’s Gemma 4 26B and 31B models now available as NVFP4 checkpoints for even faster performance on NVIDIA Blackwell GPUs. Pair the NVFP4 checkpoints with Google’s new Multi-Token Prediction drafters to get up to 3x faster inference at identical output quality, enabling frontier-class reasoning to run locally on NVIDIA GPUs. 

Mistral Medium version 3.5, also released in April, includes compatibility updates with llama.cpp and Ollama, enabling users to run on NVIDIA RTX PRO and DGX Spark systems. 

NVIDIA recently introduced NVIDIA NemoClaw, an open source stack that optimizes OpenClaw experiences on NVIDIA devices by increasing security and supporting local models. NemoClaw now supports Windows Subsystem for Linux (WSL2), bringing support to enthusiasts and developers on Microsoft’s platform. Get started with NemoClaw on DGX Spark with this step-by-step playbook.

Plug in to NVIDIA AI PC on Facebook, Instagram, TikTok and X — and stay informed by subscribing to the RTX AI PC newsletter.

Follow NVIDIA Workstation on LinkedIn and X. 

See notice regarding software product information.

From finance and procurement to supply chain and manufacturing, specialized AI agents are moving into the enterprise systems where business decisions are made, data is accessed and workflows run at scale. 

Announced today at SAP Sapphire — where NVIDIA founder and CEO Jensen Huang joined SAP CEO Christian Klein’s keynote by video — SAP and NVIDIA’s expanded collaboration helps enterprises run specialized agents with security and governance controls. 

SAP embeds NVIDIA OpenShell — an open source runtime for securely developing and deploying autonomous AI agents — into SAP Business AI Platform. In addition, SAP engineers are codesigning OpenShell alongside NVIDIA, contributing back to the open source project.

OpenShell provides isolated execution environments, policy enforcement at the filesystem and network layers, and infrastructure-level containment that guards against damage when agent logic fails. 

Within SAP Business AI Platform, OpenShell is the runtime security layer for all SAP AI agents, including custom agents built in Joule Studio — SAP’s environment for building and managing end-to-end enterprise agents.

For enterprises, the shift from AI assistants to autonomous agents changes the trust equation. An agent that can touch systems of record, cross application boundaries and operate without review at every step needs boundaries, policy enforcement and an audit trail before it can become part of production work. That’s what SAP and NVIDIA are collaborating to address.

Why the Application Layer Matters

Huang has described AI as a five-layer cake: energy, chips, infrastructure, models and applications. 

Applications sit at the top, where AI creates economic value and drives productivity for knowledge workers. As a global leader in enterprise applications and business AI, SAP is an important catalyst in the application layer, running finance, procurement, supply chain and manufacturing workflows where agents must operate within policy, identity and process controls.

That makes SAP’s position at the core of enterprise operations a key driver for enterprise adoption of agentic AI. 

Business agents need to understand roles, processes, permissions and data boundaries. They also need an execution environment that limits what an agent sees, what it can do and where inference runs.

Codeveloping the Agentic AI Foundation

NVIDIA brings its own perspective as a longstanding SAP customer — running finance, supply chain and logistics on SAP, giving both companies shared context for what enterprise-grade governance requires in practice.

SAP engineers are working alongside NVIDIA’s to further develop OpenShell’s open source codebase, focusing contributions on what enterprises need to run agentic AI in production: runtime hardening, policy modeling, enterprise identity integration, and auditing and governance hooks.

SAP and NVIDIA technologies work in unison to address important enterprise requirements for deploying trustworthy agents. 

• NVIDIA OpenShell asks: Can this agent action safely execute? 
• Joule Studio runtime — the enterprise control layer within SAP Business AI Platform — asks: Should this action happen at all? 

Together, they close a gap that application-layer security alone cannot.

A Faster Start for Agent Builders

SAP customers building custom agents will have a faster path to production. NVIDIA NemoClaw — a reference blueprint for developing and deploying autonomous agents — will be available directly in Joule Studio. This means development teams get a structured route from initial build to trusted production deployment, without having to engineer security scaffolding from scratch.

AI agents will create value only when enterprises can trust them with their data. 

For many organizations, that data is in SAP — which houses the systems of record across finance, procurement and supply chain that run their business. 

Together, SAP and NVIDIA are making agents ready to act, while staying within the boundaries enterprises require.

Learn more about NVIDIA OpenShell and NemoClaw.

See notice regarding software product information. 

“You are entering the world at an extraordinary moment,” NVIDIA founder and CEO Jensen Huang told graduates as he delivered the keynote address at Carnegie Mellon University’s 128th commencement ceremony on Sunday. “A new industry is being born. A new era of science and discovery is beginning.”

“No generation has entered the world with more powerful tools — or greater opportunities — than you,” said Huang, addressing the assembled thousands on a rainy morning at Gesling Stadium on the university’s main campus in Pittsburgh, Pennsylvania. “We are all standing at the same starting line. This is your moment to help shape what comes next.”

After encouraging graduates to turn to their mothers and wish them a happy Mother’s Day, Huang drew a direct parallel between starting his career at the beginning of the PC revolution and graduates starting theirs at the beginning of the AI revolution, emphasizing that every major computing platform shift — PCs, the internet, mobile and cloud — had led to this shared moment.

“But what is about to happen now is bigger than anything before,” he said. “Because intelligence is foundational to every industry, every industry will change.”

As a result, no graduating class is better primed than the present one to press the advantage.

“For the first time, the power of computing and intelligence can truly reach everyone and close the technology divide,” Huang said. “Now it’s your time to realize your dreams — and the timing could not be more perfect.”

AI Gives America the Opportunity to Build Again

Huang described AI as driving the largest technology infrastructure buildout in human history, and a “once-in-a-generation opportunity to reindustrialize America and restore the nation’s capacity to build.”

The American dream of opportunity — and promise of reinvention — underpins the AI revolution and its far-reaching impacts on American industry and society.

Huang underscored that AI is making intelligence more broadly accessible — reaffirming the imperative for AI to reach everyone, not just a select few. Its opportunity extends across many industries and jobs including electricians, plumbers, ironworkers, technicians and all kinds of builders.

“This is your time,” Huang said. “AI is not just creating a new computing industry. It is creating a new industrial era.”

Massive industrial and economic shifts always bring with them uncertainty, the AI revolution is no different.

“Every major technological revolution in history created fear alongside opportunity,” Huang said. “When society engages technology openly, responsibly and optimistically, we expand human potential far more than we diminish it.”

Huang explained that AI automates tasks but elevates workers. The task and purpose of a job are not the same. Radiologists, for example, don’t just read scans — they care for patients. AI automates scan reading (the task) but elevates the radiologist: the purpose. The way forward for this generation, indeed for everyone, is to engage deeply with AI, he said.

Advancing AI Responsibly

Huang emphasized that such a massive effort requires a “clear-eyed” approach if AI’s “great promise” is to be realized while addressing “real risks.”

“The responsibility of our generation is not only to advance AI — but to advance it wisely,” he said, striking a chord with the commencement crowd, who cheered and applauded his next remark that “scientists and engineers have a profound responsibility to advance AI capabilities and AI safety together.” 

And Huang didn’t leave out non-technical groups.

“Policymakers have a responsibility to create thoughtful guardrails that protect society while still allowing innovation, discovery and progress to move forward,” he said.

To meet the moment of the AI revolution, Huang counseled doing four things at once: “Advance safely. Create thoughtful policies. Make AI broadly accessible. And encourage everyone to engage.”

“History shows that societies that retreat from technology do not stop progress — they only surrender the opportunity to shape it and to benefit from it,” he said. “So, the answer is not to fear the future. The answer is to guide it wisely, build it responsibly and ensure that its benefits reach as many people as possible.”

Carnegie Mellon’s Role in Enabling the AI Era

Looking at the graduation class and their loved ones, Huang saw a part of himself and his own path in the United States.

“Like many in this audience, I am a first-generation immigrant,” Huang said, describing his view of America growing up as “not easy, but full of opportunities. Not a guarantee, but a chance.” 

“My parents came here because they believed America could give their children a chance,” Huang said. “How can we not be romantic about America?”

And in addressing the graduates and the nation’s future, Huang also acknowledged the past — namely Carnegie Mellon University’s role as “one of the true birthplaces of artificial intelligence and robotics.”

“AI started right here at Carnegie Mellon,” said Huang, invoking a string of watershed moments, from CMU researchers’ creation of the Logic Theorist in the 1950s — widely considered the first AI computer program — to the foundation of the Robotics Institute, the first academic institute devoted entirely to robotics, in 1979.

Huang also received an Honorary Doctor of Science and Technology, one of the university’s highest distinctions, from CMU President Farnam Jahanian. 

Ahead of the commencement ceremony, Huang visited Carnegie Mellon’s Robotics Institute, where he met with students — including members of the robotics club — to learn about their work and how it addresses real-world challenges.

“Carnegie Mellon has a motto I love: ‘My heart is in the work.’ So put your heart in the work,” Huang said to conclude his address. “Build something worthy of your education, your potential and the people who believed in you long before the world did.”

Watch the full address below: 

AI will help build the energy it needs.

That’s the case U.S. Energy Secretary Chris Wright and NVIDIA Vice President of Hyperscale and High-Performance Computing Ian Buck made Thursday morning at the SCSP AI+ Expo. The 30-minute fireside chat, moderated by SCSP president Ylli Bajraktari, was called “Powering the Next American Century.”

Their argument: American leadership in AI runs through American leadership in energy.

“Energy is life,” Wright said. “The more energy you have, the more affordable energy you have, the more opportunities you have in your society.”

“

The more energy you have, the more affordable energy you have, the more opportunities you have in your society.

Chris Wright, U.S. Energy Secretary — at the SCSP AI+ Expo

The Genesis Mission — the U.S. Department of Energy (DOE)’s effort to apply AI to scientific discovery — is where that case meets execution. NVIDIA is among the DOE partners on the mission, building on what Buck called two decades of NVIDIA building supercomputers with the national labs.

“NVIDIA is 100% committed and invested in Genesis,” Buck said. “I’ve never seen more excitement across the lab and industry.”

The session was one of several SCSP panels this week with NVIDIA leaders on stage: Cofounder and NVIDIA Fellow Chris Malachowsky will lead a panel on the AI+ Careers Workforce Task Force; Rev Lebaredian will speak on physical AI and simulation; Dion Harris will discuss AI-accelerated American science and AI infrastructure for Africa; and John Josephakis will join a session on U.S. quantum leadership.

The DOE Partnership

The DOE brings 17 national labs, the scientists, the national problems and the data. NVIDIA brings the full stack — not just chips, Buck said, but algorithms, methods and 20 years of partnership with the labs.

The work is happening at scale. NVIDIA and the DOE are building two AI supercomputers together at Argonne National Laboratory. The first, Equinox, is being stood up now with 10,000 NVIDIA Grace Blackwell GPUs — what Buck called “the same GPU, the same software being used to train and build AI that we’re all enjoying today.” The second, Solstice, will use 100,000 GPUs with NVIDIA Vera Rubin.

“To put that 100,000 in perspective on the next-generation GPU, which is dedicated to science, it’s 5,000 exaflops,” Buck said. “That’s a big number that actually is five times larger than the entire TOP500 supercomputer list combined.”

“We’re creating all the same technology, all the same hardware, all the same software building blocks used by all the major AI labs around the world,” Buck said, “for all of world science to go get access to.”

What that looks like in practice: Buck described an open source NVIDIA AI model trained on 1.5 million physics papers, then fine-tuned on 100,000 papers specifically about fusion. The result is a specialized AI agent DOE researchers can interrogate to advance their work faster. 

Energy and the Pace of Building

Over the last 20 years, Wright said, the U.S. has tripled oil production and doubled natural gas production — but barely grown electricity production. That’s a problem because, as Wright put it, the most important source of energy for AI is electricity.

His department is leaning back into all three pillars of the U.S. grid: natural gas, nuclear and coal.

On nuclear, Wright pointed to small modular reactors as a near-term lever — three small modular reactors (SMRs) will go critical by July 4 of this year, he said, with both new large reactors and additional SMRs to follow. 

On fusion, his department has stood up a strategic fusion office, and the lab and university programs are being, in his words, “hypercharged” by the computing power and insights AI now provides.

“We have to fix this bureaucratic and complex electricity grid so that it can grow fast, so that it can grow like our primary energy production and it can keep up with AI,” Wright said. “If we don’t do that, we’re going to slow down AI.”

NVIDIA founder and CEO Jensen Huang has described AI as a five-layer cake: energy, chips, infrastructure, models and applications. Wright’s department covers the bottom layer.

Buck took up the next layer. Asked about the intersection of energy and AI, he pointed to per-watt efficiency gains in NVIDIA chips with each generation.

“We went from the Hopper generation to Blackwell,” Buck said. “We increased performance by 30x. We actually increased performance per watt by 25 times.”

Wright returned to the grid. AI, he said, can break the bottleneck of grid interconnection studies that today take years.

“With AI, we’re going to take something that was years long and make it weeks or hours,” Wright said.

With AI, we’re going to take something that was years long and make it weeks or hours.

Chris Wright, U.S. Energy Secretary

What Success Looks Like

Asked what success looks like 12 months in, Wright pointed to fusion, materials and grid interconnection — concrete deliverables.

“We will have deliverables that we’re going to point to — we couldn’t do that before, and now we can,” Wright said. “That’s the goal of Genesis: drive discovery and bring the benefits to humans.”

There’s growing public concern that AI and data centers will drive up electricity costs, he noted. The reality, Wright said, runs the other way: “Building more electrical generation, building data centers, are actually the mechanism to lower the cost of electricity in our country and make our grid stronger.”

AI and energy are both key to human progress.

“AI doesn’t love, it doesn’t have passion,” Wright said. “It’s just going to make humans more powerful and better at pursuing whatever your passions are. It’s a thing that supercharges humans — it does not replace you.”

“

It’s a thing that supercharges humans — it does not replace you.

Chris Wright, U.S. Energy Secretary

Less typing, more tanking.

Faster logins mean more time in the gaming action — and this week provides GeForce NOW members with a smoother path straight into the battlefield.

Cloud gaming is all about instant access to titles across devices, and the latest GeForce NOW update removes another layer for members jumping into their Gaijin libraries from anywhere.

With seven games joining the cloud this week, the action keeps growing — now backed by NVIDIA GeForce RTX 5080 performance for Ultimate members across nearly the full GeForce NOW library, spanning both Ready-to-Play and Install-to-Play titles.

One Gaijin Login to Rule Them All

Gaijin account linking is now cleared for takeoff on GeForce NOW, making it easier to jump straight into battle without extra logins. With Gaijin single sign-on, one quick sign-in connects a Gaijin.net account to the cloud, so supported titles are ready to launch with fewer clicks and no password juggling.

Just head to the GeForce NOW app on PC or Mac, dive into the settings and log in under “Connections.”

Once linked, supported Gaijin games on GeForce NOW — like the combined-arms chaos of War Thunder — are ready to stream without logging in each time. It’s the same library-first, platform-linking magic members already enjoy with Xbox, Ubisoft Connect and more, now extended to Gaijin so existing PC accounts work seamlessly across devices.

Groovy New Games

Dead as Disco from Brain Jar Games drops players into a neon-soaked fever dream where every punch, dodge and takedown hits on the beat. This rhythm-infused brawler blends stylish combat with pulsing tracks, turning each fight into a dance floor showdown. Chain combos, stay in sync and keep the groove alive as the chaos escalates. Now available to stream on GeForce NOW, it’s all about timing, flair and fighting like the music never stops

In addition, members can look for the following games, all ready to play on GeForce RTX 5080-power for Ultimate members:

• Dead as Disco (New release on Steam, May 5)
• HUNTDOWN: OVERTIME (New release on Steam, May 7)
• Nuclear Option (Steam)
• Sintopia (Steam)
• Kiln (Steam and Xbox, available on Game Pass)
• Hotel Architect (Steam)
• PowerWash Simulator 2 (Xbox, available on Game Pass)

What are you planning to play this weekend? Let us know on X or in the comments below.

What has been your most-played game so far this year?

— NVIDIA GeForce NOW (@NVIDIAGFN) May 5, 2026

The race to build the world’s most powerful AI factories demands networking that keeps pace with the ambitions of AI itself. 

NVIDIA Spectrum-X Ethernet scale-out infrastructure stands at the forefront of that race as the most advanced AI networking technology available today, deployed by industry leaders who can’t afford to compromise on performance, resilience or scale.

That includes OpenAI, Microsoft and Oracle.

Companies including NVIDIA, Microsoft and OpenAI have demonstrated industry leadership by introducing Multipath Reliable Connection (MRC), an RDMA transport protocol. MRC enables a single RDMA connection to distribute traffic across multiple network paths, improving throughput, load balancing and availability for large-scale AI training fabrics. 

Think of it as replacing a single-lane road spanning a town with a cleverly laid-out street grid system paired with an on-the-fly traffic app, enabling drivers to reroute around slowdowns and road closures.

“Deploying MRC in the Blackwell generation was very successful and was made possible by a strong collaboration with NVIDIA,” said Sachin Katti, head of industrial compute at OpenAI. “MRC’s end-to-end approach enabled us to avoid much of the typical network-related slowdowns and interruptions and maintain the efficiency of frontier training runs at scale.”

In addition, Microsoft and NVIDIA have a longstanding collaboration focused on advancing the infrastructure required for the next generation of AI. Microsoft’s Fairwater and Oracle Cloud Infrastructure (OCI’s) Abilene data center, two of the largest AI factories purpose-built for training and deploying leading-edge frontier LLMs, rely on MRC to deliver on performance, scale and efficiency requirements. NVIDIA Spectrum-X Ethernet is suited for this environment, helping provide the network foundation needed to run large-scale AI models and applications with confidence.

Proven first in production with performance optimized on NVIDIA Spectrum-X Ethernet hardware and now released as an open specification through the Open Compute Project, MRC demonstrates the power of the Spectrum-X Ethernet platform: purpose-built hardware, deep telemetry and intelligent fabric control working together to take a new protocol — a set of rules that controls how data moves between two systems across a network — from concept to gigascale AI production.

MRC delivers high levels of GPU utilization by load-balancing traffic across all available paths, enabling every GPU to get the bandwidth it needs throughout a training run. It sustains high bandwidth even under congestion by dynamically avoiding overloaded paths in real time.

When data loss occurs, intelligent retransmission enables rapid, precise recovery, minimizing the impact of short-lived interruptions to long-running jobs, helping avoid GPU idle time.

Administrators also gain fine-grained visibility and control over traffic paths, simplifying operations and accelerating troubleshooting at scale.

MRC, deployed on Spectrum-X Ethernet, is optimized and engineered for resilience at massive scale. Its failure bypass technology can — in just microseconds — detect a network path failure and reroute traffic automatically in hardware. 

This failure bypass technology matters for AI training clusters where thousands of GPUs must stay synchronized, as even a brief network disruption can slow or interrupt an entire training job. Spectrum-X Ethernet prevents that by responding at hardware speed, keeping traffic flowing along precise pathways across gigascale AI fabrics.

Another innovation key to achieving gigascale AI factories is multiplanar network designs, which OpenAI deploys with Spectrum-X Ethernet in conjunction with MRC. A multiplane network consists of multiple independent network fabrics, or planes, with each providing an alternate communication path between GPUs. 

The NVIDIA Spectrum-X Multiplane capability enhances this network architecture by supporting hardware-accelerated load balancing across the planes, boosting resiliency and scale without sacrificing performance. This keeps latencies predictably low while scaling to hundreds of thousands of GPUs.   

With Spectrum-X Ethernet, customers are provided with a choice of RDMA transport models. Both Spectrum-X Ethernet Adaptive RDMA and MRC protocols, as well as other custom protocols, run natively across NVIDIA ConnectX SuperNICs and Spectrum-X Ethernet switches and support multiplanar network designs at gigascale. 

In this way, the Spectrum-X Ethernet hardware and software infrastructure that powers today’s largest AI clusters gives customers the flexibility to choose the right transport for their workload.

The MRC transport protocol is the latest example of how the industry is using Spectrum-X Ethernet as a flexible, composable platform that integrates across the full breadth of modern AI infrastructure. 

As AI factories continue to scale, the network must do more than move data quickly. It must be intelligent, resilient and based on open standards. NVIDIA Spectrum-X Ethernet delivers on all three, and with MRC, it continues to set the standard for advanced AI networking.

NVIDIA collaborated on MRC development with AMD, Broadcom, Intel, Microsoft and OpenAI.

Learn more about NVIDIA Spectrum-X Ethernet on the webpage, datasheet and technical whitepaper. 

See notice regarding software product information. 

Enterprise AI has learned to generate. It has learned to reason. Now companies are asking the next question: How should AI act?

Early agent systems have shown what’s possible, moving beyond simple prompts to take on more complex tasks. The next step is bringing those capabilities into enterprise environments — where agents must operate with context, control and consistency across real workflows.

At ServiceNow Knowledge 2026, NVIDIA founder and CEO Jensen Huang joined ServiceNow chairman and CEO Bill McDermott during the opening keynote to discuss the next phase of enterprise AI. 

The companies are expanding their collaboration across the full stack, delivering specialized autonomous AI agents that are safe and easy to adopt — powered by NVIDIA accelerated computing, open models, domain-specific skills and secure agent execution software, and bringing together enterprise workflow context from ServiceNow Action Fabric and governance from ServiceNow AI Control Tower.

ServiceNow is introducing Project Arc, a long-running, self-evolving autonomous desktop agent designed for knowledge workers, including developers, IT teams and administrators. 

Unlike standalone AI agents, Project Arc connects natively to the ServiceNow AI Platform through ServiceNow Action Fabric to bring governance, auditability and workflow intelligence to every action the autonomous desktop agent takes. It can access the local file systems, terminals and applications installed on a machine to complete complex, multistep tasks that traditional automation can’t handle, but with the controls enterprises actually need to deploy AI at scale.

The work is designed based on three requirements every company will need for long-running, autonomous agents: open models and domain-specific skills that can be customized and security that helps agents act without exposing sensitive data or systems — all running on AI factories that deliver efficient tokenomics.

Bringing this level of autonomy to enterprises requires control from the start.

Project Arc uses NVIDIA OpenShell, an open source secure runtime for developing and deploying autonomous agents in sandboxed, policy-governed environments. ServiceNow is building on and contributing to OpenShell to advance a common foundation for secure, enterprise-grade agent execution. With OpenShell, enterprises can define what an agent can see, which tools it can use and how each action is contained. 

“Project Arc represents the next step in our ongoing collaboration with NVIDIA, bringing autonomous execution to the desktop,” said Jon Sigler, executive vice president and general manager of AI Platform at ServiceNow. “By combining OpenShell’s runtime layer with ServiceNow AI Control Tower, and powered by ServiceNow Action Fabric, we’re delivering the governance and security that enterprise AI requires.” 

Open Models and Agent Skills Scale Enterprise AI

To be effective, enterprise AI systems must be adaptable. NVIDIA and ServiceNow are building on an open ecosystem that allows organizations to tailor models and applications to their specific domains and data.

NVIDIA agent skills enable specialized agents, such as ServiceNow AI Specialists, to deliver targeted capabilities across enterprise workflows. For example, the NVIDIA AI-Q Blueprint for building specialized deep research agents empowers ServiceNow AI Specialists to gather context, synthesize information and support more complex decision-making across business functions. 

In addition, the NVIDIA Agent Toolkit, including NVIDIA Nemotron open models, provide flexible building blocks and specialized skills for developing customized AI applications. To support real-world performance that these systems can perform reliably, the companies are also advancing NOWAI-Bench, an open benchmarking suite for enterprise AI agents, integrated with the NVIDIA NeMo Gym library. NOWAI-Bench includes EnterpriseOps-Gym, one of the industry’s most challenging enterprise agent benchmarks, where Nemotron 3 Super currently ranks No. 1 among open source models.

Unlike general benchmarks, these evaluations focus on multistep workflows — where enterprise AI systems often encounter real challenges — helping teams build agents that perform reliably in production environments.

Efficient AI Factories

As AI agents become long running and always on, scaling them across millions of workflows requires not just capability but efficiency — making token economics central to enterprise AI.

NVIDIA AI factories are built to deliver the lowest-cost, most-efficient tokenomics for production AI. The NVIDIA Blackwell platform delivers more than 50x greater token output per watt than NVIDIA Hopper, resulting in nearly 35x lower cost per million tokens. For enterprises running agents across millions of workflows, that efficiency can determine how quickly AI moves from pilots to broad production use.

ServiceNow AI Control Tower integrates with the NVIDIA Enterprise AI Factory validated design, extending governance and observability to large-scale AI workloads. With added agent observability capabilities, organizations can monitor behavior in real time and manage AI systems across their full lifecycle — from deployment to optimization.

AI is becoming a new way that work gets done. What’s changing now is that the core pieces required to deploy it at scale — capable agents, built-in guardrails and proven performance — are all coming together.

The companies that move fastest will be the ones that give agents the infrastructure to act, the context to make decisions and the governance to keep every action accountable — and NVIDIA and ServiceNow are making this a reality for the world’s enterprises.

Learn more about NVIDIA OpenShell and the NVIDIA AI-Q Blueprint. 

Editor’s note: This post is part of the Nemotron Labs blog series, which explores how the latest open models, datasets and training techniques help businesses build specialized AI systems and applications on NVIDIA platforms. Each post highlights practical ways to use an open stack to deliver real value in production — from transparent research copilots to scalable AI agents.

By early 2026, the open source project OpenClaw had become a phenomenon. In January, its GitHub star count crossed 100,000 as developer interest surged. Community dashboards and traffic analytics showed more than 2 million visitors in a single week. By March, OpenClaw topped 250,000 stars — overtaking React to become the most-starred software project on GitHub in just 60 days.

Created by Peter Steinberger, OpenClaw is a self-hosted, persistent AI assistant designed to run locally or on private servers. The project drew attention for its accessibility and unbounded autonomy: Users could deploy an AI model locally without depending on cloud infrastructure or external application programming interfaces (APIs).

Most AI agents today are triggered by a prompt, complete a defined task and then stop running. A long-running autonomous agent, or “claw,” works differently. These agents run persistently in the background, completing tasks on their own and surfacing only what requires a human decision. They operate on a heartbeat: At regular intervals, they check their task list, evaluate what needs action, and either act or wait for the next cycle.

OpenClaw’s rapid adoption also sparked debate. Security researchers raised concerns about how self-hosted AI tools manage sensitive data, authentication and model updates. Others questioned whether local deployments could expose users to new risks — from unpatched server instances to malicious contributions in community forks. As contributors and maintainers worked to address these issues, OpenClaw’s rise prompted a broader conversation across the AI ecosystem about the trade-offs between openness, privacy and safety.

To help enhance the security and robustness of the OpenClaw project, NVIDIA is collaborating with Steinberger and the OpenClaw developer community to address potential vulnerabilities, as detailed in a recent blog post by OpenClaw.

NVIDIA contributes code and guidance focused on improving model isolation, better managing local data access and strengthening the processes for verifying community code contributions. The goal is to support the project’s momentum by contributing its security and systems expertise in an open, transparent way that strengthens the community’s work while preserving OpenClaw’s independent governance.

 To help make long-running agents safer for enterprises, NVIDIA also introduced NVIDIA NemoClaw, a reference implementation that uses a single command to install OpenClaw, the NVIDIA OpenShell secure runtime and NVIDIA Nemotron open models with hardened defaults for networking, data access and security. NemoClaw serves as a blueprint for organizations to deploy claws more securely.

Inference Demand Multiplies With Each AI Wave

AI has moved through four phases, and the time between each is shortening. Predictive AI took years to become mainstream. Generative AI moved faster. Reasoning AI arrived faster still. Autonomous AI — the wave OpenClaw represents — is setting an even faster pace.

What compounds with each wave is inference demand. Generative AI increased token usage over predictive AI. Reasoning AI increased it another 100x. Autonomous agents, which run continuously and act across long time horizons, drive inference demand up by another 1,000x over reasoning AI. Each wave multiplies the compute required.

This increase in token usage is enabling organizations to speed their productivity by orders of magnitude. For example, long-running agents can help researchers work through a problem overnight, iterate on a design across thousands of configurations, or monitor systems and surface only the anomalies that require human judgment — freeing up researchers’ work days for higher-value tasks.

Choosing the Tool: When to Deploy a ‘Claw’

While generative AI has become a staple for on-demand tasks, there are specific scenarios where the persistent “heartbeat” of a claw offers distinct advantages. Determining when to move from a standard prompt-based AI to a long-running agent often comes down to the nature of the workflow:

• From “On-Demand” to “Always-On”: While standard models are excellent for immediate, human-triggered queries, claws are often better suited for tasks that require continuous background monitoring or periodic system checks without a manual start.
• Managing High-Iteration Loops: For complex problems, like testing thousands of chemical combinations or simulating infrastructure stress tests, a claw can manage the sheer volume of iterations that might otherwise be bottlenecked by human intervention.
• Shifting from Suggestions to Actions: In many workflows, standard AI is used to provide information or drafts. A claw is often considered when the goal is for the AI to move into the execution phase — interacting with APIs, updating databases or managing files across a long time horizon.
• Resource Optimization: For massive, token-heavy reasoning tasks, deploying a local claw on dedicated hardware like an NVIDIA DGX Spark personal AI supercomputer allows for more predictable costs and data privacy compared with high-frequency cloud API calls.

How Are Organizations Using Long-Running Autonomous Agents?

The practical applications of long-running autonomous agents span every function and sector.

In financial services, agents continuously monitor trading systems and regulatory feeds, flagging material events before the morning review. In drug discovery, agents sweep new scientific literature, extracting relevant findings and updating internal databases in real time without researcher intervention — a process that previously took weeks.

In engineering and manufacturing, agents speed problem analysis by testing thousands of parameter combinations, ranking results and flagging the configurations worth examining — and all this can happen overnight. 

In IT operations, agents diagnose infrastructure incidents, apply known remediations and escalate only the novel problems — compressing average time to resolution from hours to minutes. At ServiceNow, AI specialists leveraging Apriel and NVIDIA Nemotron models can resolve 90% of tickets autonomously. 

How Can Companies Deploy Autonomous Agents Responsibly? 

Autonomous agents are hands-on. They can send communications, write files, call APIs and update live systems. When an agent produces a wrong action, there are real consequences. Getting the accountability framework right from the start is essential, and organizations deploying autonomous agents in production must treat governance as a first-order requirement.

Organizations need to see what their agents are doing, inspect their reasoning at each step, audit their actions and intervene when needed. 

Organizations deploying autonomous agents responsibly are focused on three priorities: 

• An open, auditable framework: NemoClaw is built on OpenClaw’s MIT licensed codebase, which means organizations own the full agent harness. They can read, fork and modify every layer of how their agents are built and deployed. That transparency enables teams to understand and control the system at the code level. Running open source models like NVIDIA Nemotron locally keeps sensitive workloads, including patient records, legal documents, financial transactions and proprietary research, within the organization’s own environment, ensuring that trace data stays under organizational control.
• Securing the runtime environment: NemoClaw runs agents inside OpenShell, a sandboxed environment that defines precisely what the agent can and cannot do, enforcing clear permission boundaries from the start. 
• Local compute: NVIDIA DGX Spark supercomputers deliver data-center-class GPU performance in a deskside form factor built for continuous local inference that’s always on, with local model hosting and data that stays within the organization’s environment. NVIDIA DGX Station systems scale that capability for teams running multiple agents simultaneously across complex, sustained workloads. 

The organizations defining what autonomous agents do in practice are accumulating something valuable: months of live operational learning, governance frameworks developed through real workloads and agents that have absorbed the institutional context that makes them genuinely useful. This foundation will only deepen over time.

Get Started With NVIDIA NemoClaw

Access a step-by-step tutorial on how to build a more secure AI agent with NemoClaw on NVIDIA DGX Spark. Explore how NemoClaw can deploy more secure, always-on AI assistants with a single command.​ 

Experiment with NemoClaw, available on GitHub, and join the community of developers on Discord building with NemoClaw using NVIDIA Nemotron 3 Super and Telegram on DGX Spark.

Stay up to date on agentic AI, NVIDIA Nemotron and more by subscribing to NVIDIA AI news, joining the community and following NVIDIA AI on LinkedIn, Instagram, X and Facebook.  

Explore self-paced video tutorials and livestreams.

[Editor’s note] The blog has been updated to note that GeForce RTX 5080-power expansion also extends to the Install-to-Play library.

It’s gonna be May — and the cloud’s in full festival mode. 

16 games are joining GeForce NOW this month, including new AAA titles arriving on launch day from Steam, Xbox, PC Game Pass and GOG, so members can stream their PC libraries instantly across almost any device. Headline launches include the highly anticipated Forza Horizon 6 and 007 First Light, both ready to hit the cloud on day one.

Meanwhile, the GeForce NOW Ultimate membership dials things up with expanded RTX 5080‑class performance across the cloud gaming library, bringing higher frame rates, richer visuals and more responsive gameplay to even more of the games members already love. Ultimate members now get priority access to RTX 5080‑class rigs, making it easier than ever to tap into next‑generation PC power from almost any device.

GeForce NOW is also celebrating 30 years of Firaxis in style, with more of the studio’s classics coming as Install‑to‑Play titles for instant access — just in time for a Steam celebration sale on select Firaxis games supported in the cloud.

And this week kicks things off with six new games dropping into the cloud, setting the stage for a May packed with new adventures.

The Cloud Keeps Getting Bigger

AAA titles are landing throughout the month on their launch dates. Members can pick them up from their favorite PC stores — Steam, Xbox, PC Game Pass and GOG — and stream their PC gaming libraries instantly on GeForce NOW without having to repurchase titles to play across devices.

The highly anticipated titles hitting the cloud this month include: 

Forza Horizon 6 from Playground Games: The Horizon festival is back and bigger than ever. This is Horizon Japan — stretching from the iconic downtown streets of Tokyo City all the way to the snowy Japanese Alps — introducing the most dense and vertical map yet. A massive new open world, a stacked car roster and that legendary “one more race” energy are all heading to GeForce NOW on the title’s launch date. Members can pick it up on PC via Steam and Xbox, including Game Pass, then hit full throttle from the cloud on almost any device.

007 First Light from IO Interactive: The game brings a modern James Bond origin story to the cloud when it launches Wednesday, May 27, for PC, inviting members to step into the shadows as Bond begins his journey. Experience a mix of stealth and action built around a “breathing” gameplay loop with IOI’s cinematic, set piece-driven storytelling. Approach encounters through stealth, direct action or creative improvisation, with multiple viable paths to each objective that let players define Bond’s style. It’s ready on day one for players grabbing it on Steam, Xbox on PC or GOG.

More Power

GeForce NOW Ultimate is expanding what premium cloud gaming looks like. Starting today, Ultimate members can stream even more of their games on RTX 5080 virtual gaming rigs — bringing the power of the NVIDIA Blackwell RTX architecture to a wide range of titles.

This update significantly broadens access to 5080 performance beyond the list of GeForce RTX 5080-optimized titles. Now, across nearly the entire GeForce NOW library, members can experience higher frame rates, richer visuals, and more responsive gameplay by default at up to 5K 120 frames per second or 360 fps at 1080p.

With RTX 5080 in the cloud, Ultimate members unlock the same cutting-edge features available to GeForce RTX 50 Series GPU owners. That includes NVIDIA DLSS 4 technology for sharper image quality and higher performance, NVIDIA Reflex for reduced system latency and faster input response, and advanced ray tracing for more lifelike lighting and reflections. 

With RTX 5080 powering the default Ultimate experience, GeForce NOW delivers next-generation performance to more games, extending the value of a GeForce NOW membership. See more details in the knowledgebase article.

30 Years of ‘One More Turn’

GeForce NOW is celebrating three decades of Firaxis — the legendary studio behind Civilization and XCOM — with a growing lineup of some of the best strategy games on the cloud.

Throughout May, GeForce NOW is rolling out a collection of classic Firaxis games, available as Install-to-Play titles for instant access. Expanding beyond Civilization V and Civilization VI and the series’ signature “one more turn” gameplay,  GeForce NOW will showcase the studio’s broader legacy. 

Take to the stars in Sid Meier’s Civilization: Beyond Earth, lay tracks in Sid Meier’s Railroads!, defend the globe in XCOM 2, and command the skies in Ace Patrol and Ace Patrol: Pacific Skies.

Add to that the anniversary celebration on Steam with discounts of up to 90% on games in the Firaxis catalog for a limited time. It’s the perfect time for members to build their Firaxis collection in the cloud.

May-jor New Games

Heroes of Might and Magic: Olden Era is the official prequel that returns to the roots of the genre-defining, critically acclaimed turn-based strategy series. Raise grand armies and wield devastating spells to overcome foes in solo and multiplayer, set against a grim, war-torn world of uneasy alliances and hard choices. Now available to stream on GeForce NOW, every decision shapes a path defined by power, strategy and hard-earned victories

Anno 117: PAX Romana is turning up the heat with its first downloadable content (DLC), Prophecies of Ash, dropping a massive new volcanic island into the heart of the Roman empire. Build a thriving city on Cinis, juggle new production chains like obsidian crafting and new items, and try to stay on the good side of Vulcan — a new deity to worship —  while a very active volcano looks in the background. Additionally, the DLC introduces a new resource a new trader, Ceacilia, and several new specialists. Available to stream on GeForce NOW, the DLC is ready to play without waiting on big downloads or patches — just jump in, claim the new land and see whether the empire can handle a little fire and brimstone.

Check out what else is available this week:

• Global Rescue (New release on Steam, April 27)
• s&box (New release on Steam, April 28)
• Far Far West (New release on Steam, April 28)
• INDUSTRIA 2 (New release on Steam, April 29)
• Heroes of Might and Magic: Olden Era (New release on Steam and Xbox, available on Game Pass, April 30)
• Bus Bound (New release on Steam, April 30)

And look forward to the games coming throughout the month:

• Conan Exiles (New release on Steam and Epic Games Store, May 5)
• Dead as Disco (New release on Steam, May 5)
• HUNTDOWN: OVERTIME (New release on Steam, May 7)
• Outbound (New release on Steam, May 14)
• Starminer (New release on Steam, May 27)
• Deep Rock Galactic: Rogue Core (New release on Steam, May 18)
• Forza Horizon 6 (New release on Steam and Xbox, available on Game Pass, May 19)
• Luna Abyss (New release on Steam, May 21)
• ZERO PARADES (New release on Steam, May 21)
• 007 First Light (New release on Steam, Epic Games Store and Xbox, available on the Microsoft store, May 27)
• Hotel Architect (Steam)
• Kiln (Steam and Xbox, available on Game Pass)
• Nuclear Option (Steam)
• Sintopia (Steam)
• Sudden Strike 5 (Steam)
• Super Battle Golf (Steam)

Adding to April

In addition to the 10 games announced last month, 11 more joined the GeForce NOW library:

• ‘83 (Steam)
• Crimson Desert (Xbox, available to play via Xbox Play Anywhere)
• DayZ (Xbox, available on Game Pass) 
• Diablo III (Ubisoft Connect)
• MapleStory M (Steam)
• Morbid Metal (Steam)
• PRAGMATA SKETCHBOOK – Demo (Steam)
• Rayman: 30th Anniversary Edition (Steam and Ubisoft)
• Tides of Tomorrow (Steam)
• Vampire Crawlers: The Turbo Wildcard from Vampire Survivors (Xbox)
• Windrose (Steam)

Outbound didn’t make it to the cloud this month due to its launch date being moved. Stay tuned to GFN Thursdays for the latest.

What are you planning to play this weekend? 

AI agent systems today juggle separate models for vision, speech and language — losing time and context as they pass data from one model to the other.

Unveiled today, NVIDIA Nemotron 3 Nano Omni is an open multimodal model that brings these capabilities together into one system, enabling agents to deliver faster, smarter responses with advanced reasoning across video, audio, image and text. This best-in-class model gives enterprises and developers a production path for more efficient and accurate multimodal AI agents with full deployment flexibility and control. 

Nemotron 3 Nano Omni sets a new efficiency frontier for open multimodal models with leading accuracy and low cost, topping six leaderboards for complex document intelligence, and video and audio understanding.

AI and software companies already adopting Nemotron 3 Nano Omni include Aible, Applied Scientific Intelligence (ASI), Eka Care, Foxconn, H Company, Palantir and Pyler, with Dell Technologies, Docusign, Infosys, K-Dense, Lila, Oracle and Zefr evaluating the model. 

“To build useful agents, you can’t wait seconds for a model to interpret a screen,” said Gautier Cloix, CEO of H Company. “By building on Nemotron 3 Nano Omni, our agents can rapidly interpret full HD screen recordings — something that wasn’t practical before. This isn’t just a speed boost: It’s a fundamental shift in how our agents perceive and interact with digital environments in real time.”

Nemotron 3 Nano Omni Enables Faster, Leaner Multimodal Agents

Consider an AI agent for customer support processing a screen recording while analyzing uploaded call audio and checking data logs — or an agent for finance tasked with parsing PDFs, spreadsheets, charts and voice notes. Today, most agentic systems accomplish these tasks with separate models for vision, speech and language. 

This approach increases latency through repeated inference passes, fragments context across modalities, and adds cost and inaccuracies over time.

By combining vision and audio encoders within its 30B-A3B, hybrid mixture-of-experts architecture, Nemotron 3 Nano Omni eliminates the need for separate perception models, driving inference efficiency at scale. It pairs this efficiency with strong multimodal perception accuracy, enabling AI systems to achieve 9x higher throughput than other open omni models with the same interactivity. The result is lower costs and better scalability without sacrificing responsiveness or quality.

In agentic systems, Nemotron 3 Nano Omni can work alongside proprietary cloud models or other NVIDIA Nemotron open models — such as Nemotron 3 Super for high-frequency execution or Nemotron 3 Ultra for complex planning — as well as proprietary models from other providers, to power sub-agents for agentic workflows such as computer use, document intelligence and audio-video reasoning.

• Computer use agents — Nemotron 3 Nano Omni powers the perception loop for agents navigating graphical user interfaces, reasoning over onscreen content and understanding user interface state over time. H Company’s latest computer usage agent, powered by Nemotron 3 Nano Omni, uses a native input resolution of 1920×1080 pixels to achieve high-fidelity visual reasoning. In preliminary evaluations on the OSWorld benchmark, this integration showed a significant leap in navigating complex graphical interfaces and used Nemotron 3 Nano Omni’s ability to process very high-resolution images. 
• Document intelligence — Interprets documents, charts, tables, screenshots and mixed-media inputs, enabling agents to reason across visual structure and text content coherently. Critical for enterprise analysis and compliance workflows.
• Audio and video understanding — For customer service, research and monitoring workflows, Nemotron 3 Nano Omni maintains audio-video context, tying what was said, shown and documented into a single reasoning stream instead of disconnected summaries.

Open and Customizable, Deployable Anywhere

Nemotron 3 Nano Omni is released with open weights, datasets and training techniques — giving organizations full transparency and control over how the model is customized and deployed. 

Developers can use tools like NVIDIA NeMo for customization, evaluation and optimization for domain-specific use cases. Because the Nemotron family of models is open, organizations can deploy them in environments that meet regulatory, sovereignty or data localization requirements.

The Nemotron 3 family — including Nano, Super and Ultra models — has seen over 50 million downloads in the past year. Omni extends the family’s capabilities into multimodal and agentic domains. 

The model is available on Hugging Face, OpenRouter and build.nvidia.com as an NVIDIA NIM microservice and through a broad ecosystem of NVIDIA Cloud Partners, inference platforms and cloud service providers. 

Its open, lightweight architecture supports consistent deployment from local systems like NVIDIA Jetson hardware, NVIDIA DGX Spark and DGX Station to data center and cloud environments. 

Visit the NVIDIA technical blog for tutorials, cookbooks and deployment guides for Nemotron 3 Nano Omni use cases. Stay up to date on agentic AI, NVIDIA Nemotron and more by subscribing to NVIDIA news, joining the community and following NVIDIA AI on LinkedIn, Instagram, X and Facebook.  

Explore self-paced video tutorials and livestreams.

Editor’s note: This post is part of Into the Omniverse, a series focused on how developers, 3D practitioners, and enterprises can transform their workflows using the latest advances in OpenUSD and NVIDIA Omniverse.

Manufacturing’s traditional design-build-test cycle rested on a single assumption: Real-world testing was the only reliable test environment. 

That assumption is now shifting. 

Today, high-fidelity simulation produces synthetic training data accurate enough for production-grade AI. This is enabling perception systems, reasoning models and agentic workflows to excel in live factory environments.

OpenUSD has emerged as the connective standard that makes this practical, and the manufacturers building on it are already experiencing measurable results. 

SimReady: The Content Standard for Physical AI 

As physical AI becomes integral to industrial operations, manufacturers face a foundational challenge: Assets don’t travel reliably between 3D pipelines. Every time an asset moves from a computer-aided design tool to a simulation platform, physics properties, geometry and metadata are lost — forcing teams to rebuild from scratch.

SimReady is the content standard, built on OpenUSD, defining what physically accurate 3D assets must contain to work reliably across rendering, simulation and AI training pipelines. 

In addition, NVIDIA Omniverse libraries provide the physics-accurate, photorealistic simulation layer where AI models are trained and validated before deployment. 

Four Ways Manufacturers Are Putting the NVIDIA Physical AI Stack to Work

ABB Robotics Closes the Sim-to-Real Gap at 99% Accuracy

ABB Robotics has integrated NVIDIA Omniverse libraries directly into RobotStudio HyperReality, its simulation platform used by more than 60,000 engineers globally. 

The platform represents robot stations as USD files running the same firmware as their physical counterparts, making it possible to train robots, test part tolerances and validate AI models before a production line exists. 

Synthetic training variations — such as lighting conditions and geometry differences — can be generated at scale, covering scenarios that would be impractical to replicate manually. 

“We’ve managed to vertically integrate the complete technology stack and optimize it to a point where we’re now achieving 99% accuracy on the simulated version,” said Craig McDonnell, managing director of business line industries at ABB Robotics.

The downstream outcomes: up to 50% reduction in product introduction cycles, up to 80% reduction in commissioning time and a 30-40% reduction in total equipment lifecycle cost.

JLR Compresses Four Hours of Aerodynamic Simulation to One Minute

JLR applied the same simulation-first principle to vehicle aerodynamics. Engineers trained neural surrogate models on more than 20,000 wind-tunnel-correlated computational fluid dynamics simulations across the vehicle portfolio — with 95% of aero-thermal workloads now running on NVIDIA GPUs. 

The Neural Concept Design Lab — built on Omniverse and deployed at JLR — visualizes aerodynamic changes in real time as designers adjust vehicle geometry, collapsing what was a sequential design-then-simulate cycle into a continuous loop. A result that once took four hours now takes one minute. 

Tulip Brings Real-Time Factory Intelligence to Terex for Operational Gains

Once a factory goes into production, a different intelligence challenge begins — one that simulation alone can’t address. 

Tulip Interface’s Factory Playback platform demonstrates how existing infrastructure can become an intelligence layer, turning operations records into something users can actually learn from. Tulip built Factory Playback on the NVIDIA Metropolis VSS Blueprint — a reference architecture for extracting structured intelligence from factory camera feeds — connecting camera streams, machine sensor data and operational context into a unified timeline of what actually happened. 

In addition, Factory Playback uses the NVIDIA Cosmos Reason vision language model to interpret camera streams and operator behaviors in real time, running on premises on NVIDIA GPUs.

Deployed at Terex, a global industrial equipment manufacturer with over 40 plants, the system is expected to deliver a 3% increase in yield and 10% reduction in rework. 

“I am excited to see what manufacturers will do with the power of AI to augment their daily capabilities,” said Rony Kubat, cofounder and chief information officer of Tulip Interfaces. 

Getting Started

SimReady assets, Omniverse libraries and NVIDIA’s physical AI stack provide a foundation developers can adopt, extend and combine across any industrial application. Here’s how to get started:

• See how NVIDIA and partners put physical AI to work on the factory floor at Hannover Messe.
• Start building autonomous robots, digital twins and AI-powered systems with these free, self-paced courses.
• Explore NVIDIA Isaac Sim and Omniverse libraries on the NVIDIA developer portal.
• Deploy the NVIDIA Metropolis VSS Blueprint on existing camera infrastructure to gain new insights from the shop floor. 
• Explore the SimReady Foundation specification framework on GitHub.
• Browse NVIDIA Cosmos Cookbook recipes for domain-specific physical AI applications across robotics, simulation and autonomous systems.
• Access the full Omniverse developer hub.
• Join the community to connect with fellow developers and innovators who are building the future with NVIDIA technologies.