Visit additional Tabor Communication Publications
May 10, 2011
Supercomputer maker Cray has posted a modest loss for the first quarter of 2011 and downgraded its low-end revenue expectations for the year by $20 million. In a conference call with investors, Cray CEO Peter Ungaro blamed most of this on a slowdown in government funding, as countries retreat from the spending spree of the last couple of years. Despite that, Ungaro and company are still aiming for a profitable year as they prepare to roll out new supercomputer offerings in the second half of 2011.
"Over the last quarter, we've seen a delay in government spending across most of geographies, which is having a direct impact on our business by lengthening the sales process for new systems and services," Ungaro told the investors.
In the quarter just ended, Cray's posted a $1.5 million loss on $39.9 million in revenue. While that might be disheartening after finishing 2010 with a $15.1 million profit, the good news is that revenue for first quarter grew 40 percent year from this same time last year. And the quarterly loss was nearly equal to the one-time $1.1 million payout in severance for the 52 employees that were laid off in March. "We had solid first quarter, with revenue in line or better than our guidance across the board," declared Ungaro.
For the year Cray is now projecting between $300 million and $340 in revenue for the year, about half of which they intend to close in the Q4. In general, Q1 tends to be a slow time for Cray, and the HPC business in general, so the relatively healthy $40 million start could be seen as a good omen. The decent Q1 revenue was thanks in large part to XE6 ("Baker") systems that were delivered to customers in the second half of 2010, but had not completed acceptance testing until late last year or earlier this year.
Driving the somewhat lowered revenue guidance for 2011 is the more conservative outlook for government money budgeted for federal R&D in the US and abroad. Cray is experiencing this as delays in supercomputer RFPs and contract execution. Despite that, Ungaro is still bullish on the feds, saying they have a very good win rate against IBM and other competitors in this arena, and expect a number of government supercomputer deals to close before the end of 2011. "We believe this is really about a delay, and not a reduction, in our opportunity pipeline," said Ungaro.
One area that is still performing well is Europe, where they haven't encountered any slowdown. In fact, this week Cray announced an upgrade to the University of Edinburgh XE6 supercomputer, which is part of the UK's High-End Computing Terascale Resource (HECToR) service. The new machine will get 10 additional cabinets as well as an upgrade from the existing Magny-Cours Opterons to the new 16-core "Interlagos" processors that AMD plans to launch in Q3. The Edinburgh super was upgraded last year from an XT6 to an XE6 by slipping in the new Gemini interconnect.
The worst geography for Cray right now is Japan. Due to the devastating earthquake in March, government money is being diverted into essential services and infrastructure to cope with the aftermath of the disaster. As a result, Cray doesn't expect to accrue any money from outstanding contracts until 2012. In other areas of the Asia/Pacific region, a number of other large system deployments have been delayed as well.
In the US, the company's largest market, the situation is somewhat better. The fiscal budget was recently approved, although more than six months late. Some supercomputing opportunities that were on hold are now coming back in play, said Ungaro, and his hope is that they still have enough time to win those deals, deliver the machines, and book the money before the end of the year.
One such deal Cray has in its sights is worth over $50 million and, if the supercomputing gods are smiling on the company, a good chunk of that could be realized this year. According to Ungaro, the system in question is a large XE6 machine -- undoubtedly a petascale system -- that will incorporate the upcoming Interlagos Opterons and the new NVIDIA GPU-equipped blades that Cray has been talking about. As long AMD and NVIDIA hit their milestones, Cray expects to start deploying a number of these new tricked-out XE6 supers throughout the second half of 2011.
The Interlagos and GPU upgrades for the XE6 are slated for the third and fourth quarters, respectively, and appear to be on track. As mentioned before, Interlagos (Opteron 6200 series) is due out from AMD in Q3, and as long as the chipmaker stays on schedule, Cray should be able to start shipping systems in the same quarter. Integrating the Interlagos chips should be fairly straightforward, requiring only an update to the XE6 software stack to support the new Bulldozer-architected chips.
The GPU upgrade is a little trickier in that it involves upgrades to both hardware and software. Although neither NVIDIA nor Cray are talking about it at this point, the new XE6 GPUs could very well end up being the next-generation Kepler processors, which were originally slated to launch in the second half of 2011. Given that each Kepler GPU will deliver something on the order of 1.5 teraflops, that would make these new XE6 machines even more interesting.
Perhaps the most problematic segment for Cray right now is its custom engineering (CE) business, which, because of government slowdown on the HPC services side, is predicted to slip into reverse in 2011. After booking $10 million in 2008, $30 million in 2009, and $62 million in 2010, custom engineering is now being projected to bring in just $40 to $50 million this year. "It's been on an amazing growth path since we started it in 2008." said Ungaro. "While it looks like we're taking somewhat of a breather on that growth this year, I believe were going to come out of 2011 with the ability to grow CE at a double-digit rate going forward."
But at least one CE project, the development of the second-generation XMT system, is getting set to launch. XMT is Cray's massive, multithreaded processing architecture based on the company's custom Threadstorm processor. Unlike its mainstream XE line, which is geared for conventional number-crunching scientific codes, XMT is designed for real-time analysis of unstructured data, and according to Ungaro is "unlike anything else on the market today for big data analytics."
In February, Cray announced its first XMT deal with CSCS in Switzerland. That system will be a proof point for the technology and a development platform for big data analytics applications. Cray has high hopes for the new offering going forward, especially since it gets them into a market that is poised for fast growth. For this year though, it's doubtful if the revamped XMT product will contribute a huge share to the bottom line.
With countries around the world in a more austere frame of mind, Cray certainly has its work cut out for it in 2011. The larger danger is that national governments draw back on R&D funding on a longer term basis and dry up the future pipelines for HPC infrastructure. That's unlikely to happen to any great degree, given the stated exascale goals in the US, Europe and Japan, not to mention the new competition of China. With $126 million of cash in the bank, Cray could certainly endure a temporary retraction in the business, but would much prefer if 2011 turned out to be the sequel to its success in 2010.
May 23, 2013 |
The study of climate change is one of those scientific problems where it is almost essential to model the entire Earth to attain accurate results and make worthwhile predictions. In an attempt to make climate science more accessible to smaller research facilities, NASA introduced what they call ‘Climate in a Box,’ a system they note acts as a desktop supercomputer.
May 22, 2013 |
At some point in the not-too-distant future, building powerful, miniature computing systems will be considered a hobby for high schoolers, just as robotics or even Lego-building are today. That could be made possible through recent advancements made with the Raspberry Pi computers.
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.
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.