Visit additional Tabor Communication Publications
October 15, 2008
Today Intel announced it had acquired the assets of NetEffect, an Austin-based company that makes iWARP-capable adapters. Developed as a high performance, low latency Ethernet counterpart to InfiniBand, iWARP is the RDMA over Ethernet protocol standard for cluster applications. Intel will inherit NetEffect's product portfolio, which includes 1 and 10 GbE accelerated adapters, 10 GbE adapters for blade configurations as well as a 10 GbE ASIC.
Intel says it will also hire about 30 NetEffect employees -- mostly senior engineers -- who will continue to work at the NetEffect facility in Austin, Texas. The team will report to Intel's LAN Access Division and will be incorporated into its hardware engineering group.
NetEffect filed for bankruptcy on Aug. 27 with assets of less than $1 million and debts between $10 and $50 million, according to the filing. Intel purchased those assets, which includes the Ethernet product line and the associated intellectual property, for $8 million. Why would Intel offer ten times the asset price for the company? Apparently, the deal has been in the works for awhile. On Sept. 20 The Austin American-Statesman reported that NetEffect had filed for Chapter 11 to set up the Intel acquisition:
NetEffect chief executive Rick Maule said Intel has offered $8 million for the company but insisted that NetEffect file for bankruptcy reorganization before the deal is made. Under the court's ruling, other potential buyers have until the end of September to submit bids in competition with Intel, Maule said.
Since Intel could now acquire the company under Section 383 of the bankruptcy code, they wouldn't have to deal with NetEffect's considerable debt. The $8 million dollar offering from the chipmaker made counter bids unlikely, especially considering that IT vendors are all looking down the barrel of a recession. So the one-month window of opportunity for other potential buyers came and went.
The big losers are NetEffect's private investors who anteed up $47 million between 2002 and 2006, and an additional $13.5 million this year. The investors included Austin Ventures, Duchossois Technology Partners, Granite Ventures, Infinity Capital, JatoTech Ventures, Texas Instruments and TL Ventures.
Even before today's deal, Intel was a backer of high performance Ethernet. The company is a member of the OpenFabrics Alliance, which supports RDMA-based InfiniBand and iWARP fabrics. Although once a cheerleader for InfiniBand, Intel soured on the technology when it became apparent that it would not be the unifying network fabric after all. Now with the newly-inherited iWARP hardware, Intel becomes an RDMA fabric vendor. Chelsio Communications, the remaining vendor with a deliverable iWARP solution today, now gets to compete against Intel. How fun for them.
So what's iWARP do for Intel? I spoke with Steve Schultz, director of product marketing for Intel's LAN Access Division, to get an idea of where iWARP fits into the company's overall Ethernet strategy. He says they see iWARP as the missing piece for his division's 10 GbE product portfolio. Intel divides the 10 GbE market into three application areas: virtualization, storage over Ethernet (LAN-SAN convergence), and high performance clustering. While Intel has Ethernet solutions for the first two, including FCoE and iSCSI products for storage, up until now it was missing a high performance Ethernet offering. Schultz says the latter is a niche market, but a key one for 10 GbE.
Intel believes that HPC programmers, unlike most commercial developers, are willing to write their applications to take advantage of RDMA fabrics. The HPC community is already used to proprietary fabrics or InfiniBand, so they're prepared to expend the extra effort required to get the needed performance out of their apps. "For these customers, iWARP is a perfect solution," says Schultz. "They can get all the benefits of Ethernet plus that ultra-low latency."
Of course, as I pointed out in this week's article about the new Woven Systems switch, 10 GbE for HPC is still in the wait-and-see stage, much to the frustration of companies like NetEffect and Chelsio. Since 10 GbE won't trump InfiniBand on performance, it will have to compete on price, or the more nebulous TCO. This is not likely to change until 10 GbE takes off in the enterprise, and all the commodity pieces (especially 10 GbE on the motherboard) fall into place.
Intel, unlike NetEffect, has the resources to wait this out, and perhaps even move it along should it choose to do so. One opportunity now available to the company is to drop the NetEffect 10 GbE ASIC onto an Intel server board aimed at high performance computing. Of course, that would eat into its new iWARP adapter business, but since 10 GbE will eventually end up on the motherboard anyway, why not set the pace for the rest of the industry. And since Intel makes chips for a living, it might decide to do something even more creative with the NetEffect silicon. For $8 million, Intel bought itself a lot of options.
May 16, 2013 |
When it comes to cloud, long distances mean unacceptably high latencies. Researchers from the University of Bonn in Germany examined those latency issues of doing CFD modeling in the cloud by utilizing a common CFD and its utilization in HPC instance types including both CPU and GPU cores of Amazon EC2.
May 15, 2013 |
Supercomputers at the Department of Energy’s National Energy Research Scientific Computing Center (NERSC) have worked on important computational problems such as collapse of the atomic state, the optimization of chemical catalysts, and now modeling popping bubbles.
May 10, 2013 |
Program provides cash awards up to $10,000 for the best open-source end-user applications deployed on 100G network.
May 09, 2013 |
The Japanese government has revealed its plans to best its previous K Computer efforts with what they hope will be the first exascale system...
May 08, 2013 |
For engineers looking to leverage high-performance computing, the accessibility of a cloud-based approach is a powerful draw, but there are costs that may not be readily apparent.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
In this demonstration of SGI DMF ZeroWatt disk solution, Dr. Eng Lim Goh, SGI CTO, discusses a function of SGI DMF software to reduce costs and power consumption in an exascale (Big Data) storage datacenter.
The Cray CS300-AC cluster supercomputer offers energy efficient, air-cooled design based on modular, industry-standard platforms featuring the latest processor and network technologies and a wide range of datacenter cooling requirements.