At first glance people may be surprised that we focused on applications in the access but we think it really makes sense and is a very under served market.  In order to understand why, you have to remind yourself that the access market is going to require many of the same attributes that are in today’s edge.  These include:

• – Ethernet services support (E-Line, E-LAN) (NOTE — NOT SIMPLY ETHERNET TRANSPORT!)
• – VPLS/VPWS Pseudowires
• – OAM (both CCM and performance monitoring)
• – MEF defined CIR/EIR profiles and traffic management
• – Class of service egress shaping
• – Network timing (Synch Ethernet and 1588)

From a capacity standpoint the devices are typically serving interfaces that run at no more than 1Gb but may require support for 400Mb radio interfaces or SONET services running >100Mb.  10Gb ports would be HUGE overkill and are not required at all in this or even next generation!

Oh … and it has to be low cost (~$100) and low power (~5-8 watts).

Now if you look at the technology options available to system designers they have very few options that can meet these needs.  Sure they can try and use enterprise silicon — but where is the support for E-Line services and OAM?  If they go the network processor route they blow the power and cost budget. So what can a system OEM do if they want to allow network operators to add Ethernet interfaces with “Carrier Services” into their new wireless backhaul platforms to allow these services to go directly to the cell site or demarcation point in the premise. 

We have developed the B-XLA to address this exact problem.  By adding true carrier Ethernet capabilities into a line of access ethernet switches we are solving a critical problem in the wireless backhaul and multi-tentant access demarcation platforms and providing a way for system designers to cost effectively add these capabilities.

Its an interesting market — Here are our internal estimates: 

Semiconductor Market Forecast

Lightstorm Wordle

I could not resist posting this Wordle.  I took a recent marketing piece and ran the text thru the program and it gnerated this word cloud.  Very cool representation of carrier Ethernet market and Lightstorm.  To enlarge simply click on the graphic.

“Bandwidth demands are surging due to the increased adoption of consumer and business services. The “Cisco Visual Networking Index (VNI) Forecast and Methodology, 2007-2012” forecasts IP traffic to increase sixfold between 2007 and 2012, with video representing more than half the consumer traffic. Similarly, mobile data is expected to grow by 125 percent from 2007 to 2012. To meet such dramatic increases in demand, many service providers are turning to Carrier Ethernet to significantly expand their network capacities and enable them to take advantage of this opportunity.”

Every vendor has a different approach to solving these problems and reading thru the marketing materials to figure out hype vs. feature is a challenge but clearly there is real demand in the market and today’s products require significant upgrades to achieve the required features demanded by the forthcoming traffic.

Now we have also created a 10 minute slideshow on the topic which you can just sit back and watch … or if you are one of those people that like to jump around just hit “next slide” button and the audio/slide moves ahead. 

We would welcome your thoughts and comments on this subject.

Arista is drawing a lot of attention to itself with its rockstar management line-up and pedigree (Sun Founder Andy Bechtolsheim and Cisco Alum Jayshree Ullal) but what I find refreshing is that they are also highlighting a very well known problem … ASIC development costs.

Arista announced that it is using silicon from Fulcrum Microsystems (a Lightstorm Partner) and stated the following in the WSJ article (it is subscription based so excerpt only below)

“One expense for competitors has been designing special chips. Arista, by contrast, plans to use an off-the-shelf networking chip developed by Fulcrum Microsystems Inc.,”

They continue to point out that they got to market with <50 employees and while in business for several years spent less than $20M to get their product to market.

“Force10 Networks Inc., says it has raised more than $500 million in venture capital since 1999. Mr. Bechtolsheim estimates Arista has spent only $20 million getting its products to market.”

While this is only an example of what can be done with merchant silicon it is great to see the business case for the merchant approach recieve such well publicized support.  Congratulations Fulcrum on a high profile win!

We were honored that ZTE announced that they had selected the Brooklyn-10 for use in a Video-on-Demand server for use in their highly successful IPTV portfolio.  The announcement was made yesterday at a Trade Mission in Bejing with the Taoiseach (Prime Minister) of Ireland, Mr. Brian Cowen.  

Can’t wait until the ZTE products start to enter the market and we can really demonstrate the guaranteed delivery of packets that will make the application crystal clear for customers using the IPTV product from ZTE.

Here is some press coverage of the announcement:

Irish Times

Check out this Slideshow (with voice even !) and let us know what you think!

Oh and we will post some fun stuff too that we find on the net to keep things light and without going overboard news on Lightstorm Networks as well. If you ever have ideas, comments, or just want to say Hi … drop me a note or post a comment. Wade (Wade <at> Lightstormnetworks <dot> com)