High Performance Computing Roundup
September 22nd, 2010 By: John Rath
A GPU conference, HPC conference and the upcoming SuperComputing 2010 conference supplied the industry with some news this week on innovations taking place.
HPC Financial Markets
The High Performance Computing Financial Markets show and conference was held Monday in New York and focused on speed while lowering computer costs in the financial markets. Microsoft’s Bill Hilf delivered the keynote session on how Wall Street is meeting the challenge of modeling at ever increasing speeds.
Microsoft HPC Server 2008 R2
Microsoft announced the immediate availability of Windows HPC Server 2008 R2 at the conference. “This release of Windows HPC server is a key step in our long-term goal to make the power of technical computing accessible to a broader set of customers, with capabilities across the desktop, servers and the cloud,” said Bill Hilf, general manager, Microsoft Technical Computing Group. New capabilities in HPC Server 2008 R2 a sixteen-fold increase over the past two years in application support and more powerful analysis and computation. The HPC Services for Excel 2010 example shows a calculation of a workbook with 1,700 records dropped from 14 hours previously, to less than three minutes in the new version.
Microsoft hinted that a future release of HPC Server will allow customers to provision and manage HPC nodes in Windows Azure from within on-premises server clusters. They also highlighted the simplifying of development of HPC applications for the new generation of distributed, or parallel, computing resources on client systems, server clusters and in the cloud. “Companies and products, like the new Windows HPC Server 2008 R2, help customers easily take advantage of new technology advances, such as HPC clusters, GPUs, cloud computing and multicore processors,”said Earl Joseph at IDC. “All of these enhancements will help to accelerate the growth of the high-performance computing market.”
SGI and Cray support HPC Server
HPC vendors SGI and Cray made announcements Monday in support of the new Windows HPC Server 2008 R2 release. SGI announced support on their Octane III, SGI’s personal supercomputer and that they intend to expand use of the new operating system across Rackable server and other SGI product lines in the future. “This platform will enable new computational capabilities for an extended set of customers,” said Christian Tanasescu, vice president, software engineering at SGI. “With our continuing support for HPC2008 on additional platforms, we are continuing to grow with Microsoft to provide boundless capacity in technical computing markets.”
Cray announced that the new Windows HPC Server 2008 R2 is now available on the Cray CX1 line of deskside supercomputers, including the Cray CX1 and the Cray CX1-iWS systems. “We designed the Cray CX1 systems to be easy to use, and offering a HPC product that combines the computational power of a cluster with the familiarity of the Windows HPC Server 2008 R2 operating system gives more researchers, developers and scientists access to Cray systems,” said Barry Bolding, vice president of Cray’s products division.
Nvidia GPU Conference
Nvidia is hosting a GPU Technology Conference this week in San Jose. Building on the inaugural event held last year, this year’s event covers GPU computing trends, disruptive technologies and emerging applications, tools and techniques and the opportunity to network with experts and colleagues across industries.
On Tuesday CEO Jen-Hsun Huang took the stage and for the first time in the company’s history detailed a product roadmap for future Nvidia chips. Huang boiled down the importance of performance to the metric of dual-precision flops per watt, using their Tesla GPU as the baseline.
Cray made an announcement at the conference that it is developing blades based on the Nvidia Tesla 20-Series GPUs for the Cray XE6 product line. “The Cray XE6 is designed to solve the system challenges of petascale computing just as Tesla GPUs are providing the computational power efficiency to bring this level of computing to the larger HPC community,” said Andy Keane, general manager, Tesla business at NVIDIA. “Cray’s planned Tesla GPU-based system will be a powerful combination of technologies that will provide the highest level of performance and scalability for power efficient supercomputing.”
The Nvidia Ntersect blog has a number of other announcement recaps, videos and news roundups for activities at the conference so far.
SC10 in New Orleans
The SC10 conference will take place in New Orleans November 15-18, 2010 and will focus on the technology thrusts for High Performance Computing of climate modeling, heterogenous computing and data-intensive computing. On Tuesday it was announced that Clayton M. Christensen will present the keynote speech at SC10. Christensen is a professor at the Harvard Business School and an authority on disruptive innovation as well as an advisor to numerous companies and countries.
“The HPC community is no stranger to disruptive technologies, but at this point in time, HPC faces a level of disruption like nothing we have ever faced before,” said Barry Hess, general chair of SC10 and deputy CIO at Sandia National Laboratories. “As we move forward to a world of sustained petaFLOPS and begin the decade-long transition to exascale, Disruptive Innovation and Disruptive Technologies will become commonplace terms for all of us.”
PCPosted September 22nd, 2010
Pretty funny that both SGI and Cray, two companies that are hurting badly and spiralling downward, made announcements Monday but were too ill to have a booth at the conference like all of their competitors had. Anyone can make announcements. These unproftable co’s will be gone soon.