The October 2009 list has been published in the Asian Top500 supercomputers site. With no surprise, the Shaheen supercomputer at King Abdullah University for Science and Technology takes the first position in the list, achieving 185 TeraFlops in compute power.

In the second place of October list is the notable Magic Cube supercomputer, installed at the Shanghai Supercomputer Center in China. The Magic Cube operates with 30720 cores, achieving 180 Teraflops and running Windows 2008 HPC as the operating system.

EKA system at the Computational Research Laboratories, TATA SONS, India, takes the third place with 132.8 Teraflops, running 14384 cores of Intel EM64T Xeon 3.0GHz CPUs with Infiniband DDR interconnect.

Chine is clearly the leading consumer of HPC systems with 15 of the 45 systems submitted so far. Japan has 14 systems, while India comes in the third place with 6 systems submitted.

While most of the industry has gone through an enormous consolidation over the past year, Amsterdam-based ClusterVision is defying the trend by spinning off its ClusterVisionOS HPC management software into a separate company called Bright Computing. The new business launched itself this week, along with its flagship offering, Bright Cluster Manager. The product is literally a rebranded version of a new ClusterVisionOS release that has been under development for some time.

ClusterVision is a Linux cluster integrator that builds commodity HPC systems for the European market (more broadly, the entire EMEA region). The company uses hardware from a variety of manufacturers -- white-box manufacturers, but also IBM and Dell -- and packages the clusters with its home-grown Linux-based ClusterVisionOS and software stack. As of this week, though, ClusterVisionOS is replaced by Bright Cluster Manager, which makes ClusterVision just another reseller of the software.

In fact, the newly branded software has already been shipped to European customers as part of ongoing ClusterVision deployments. Since Bright Cluster Manager is already an established product with a revenue stream in place, the company was able to launch without outside investment, and for the time being, even without a CEO.

The impetus behind spinning off CluserVision's cluster management product was the desire to make it available to other system integrators and OEMs, thus paving the way to penetrating the global market. According to Matthijs van Leeuwen, commercial director and co-founder of ClusterVision, since showcasing the latest version of ClusterVisionOS last year at SC08, they have received a good deal of encouraging feedback from other OEMs and integrators to make the product more widely available.

No reseller deals were forthcoming at Bright Computing's launch, but several are apparently in the works and will be announced in the months ahead. Bright Computing is based in San Jose, Calif., which suggests the importance of the North American market to the company, but the idea is to cover the EMEA and Asia-Pacific regions as well.

Full article: http://www.hpcwire.com/features/ClusterVision-Spins-Off-Cluster-Management-Software-Company-63702962.html

At last week's GPU Technology Conference (GTC), representatives from two of the nation's major supercomputing centers, Oak Ridge National Labs and the University of Illinois at Urbana-Champaign, indicated their plans to incorporate GPUs as key computational elements in next generation supercomputers now under construction at both sites. These organizations, along with hundreds of others, have been experimenting with the use of GPUs to boost their number crunching capabilities over the past three years, and it now appears that the scientific community is ready to fully embrace GPU computing. Almost 100 "poster papers" from researchers in academia and industry lined the halls of the conference facility, touching on topics as varied as "Nanophotonic simulation," "Accelerated path planning for multi-axis machine tools," and "Using GPUs for Internet routing processing."

Nvidia's been plugging away at these new uses for its GPUs since 2006. Fermi, the new GPU it unveiled last week, incorporates several features designed specifically to appeal to scientific users, including a far more robust 64-bit floating point capability (delivering a peak rate of 750 DP GFLOPS, compared with 80 GFLOPS in its current offering), and the use of error-correcting memory in the on-board frame buffer. It has also dramatically expanded its software capabilities, allowing developers to program its GPUs using C++, along with two new industry standard software interfaces, OpenCL and DirectCompute. Nvidia also introduced a new development tool, Nexus, that eases the problems of debugging multi-threaded applications in mixed CPU/GPU environments. There's no doubt Fermi is coming late to market (Jen-Hsun Huang, Nvidia's CEO, noted that he wished he had had it "six months ago"), but the scientists who use his chips seem willing to cut Nvidia some slack, and are ready to gobble up the new GPUs when they emerge in 2010.

AMD's also been dabbling in the GPU computing market since 2006, but Google's 2007 acquisition of a key AMD partner (PeakStream) forced a software reset in its program. AMD refocused its efforts around the new OpenCL and DirectCompute standards, but lacks the ability to run C++ and Fortran, two languages with great popularity in the high performance computing arena. AMD's new chips outperform Nvidia's forthcoming Fermi on single precision (32-bit) floating point math (2.7 TFLOPS to 1.5 TFLOPS), but lag on double-precision math (500 GFLOPS to 750 GFLOPS). AMD's design lacks support for error correcting memory; their engineers argue that they've never seen an ECC memory failure in years of hard searching for such errors, and the feature is just not needed.

While AMD and Nvidia battle for supremacy in the GPU computing market, there's one obvious loser, Intel. AMD's 5870 appeared on schedule. Although Nvidia's Fermi is late, its prior generation GTX 280 still has some life left in it. But, Intel's many-core Larrabee is still a no-show, and the company's Larrabee demo at its recent Developers' Forum was universally regarded as brain-dead, if not an outright embarrassment. End users with high performance computational requirements previously filled their data centers with racks of x86 servers to handle those requirements. Now they have now discovered and validated an alternative approach that requires fewer x86 CPUs, less power, and less space. GPU computing won't solve all the world's computing problems, but it will give users who buy their systems by the teraflop a new, more cost effective alternative that will take some of the wind out of Intel's high performance computing sales.




Note: Readers can see detailed architectural analysis of Fermi at http://tinyurl.com/ycerrxv

http://www.hpcwire.com/

Since 1993, www.top500.org has taken the initiative to assemble and maintain the list of the 500 most powerful computer systems twice a year around the world. With time, however, this list has come to be overwhelmingly populated by machines built by a few countries. While this is an accurate overall reflection of the state of supercomputing in the world, it makes more difficult an understanding of the relative computing capability within Asia and the Middle East.

To provide this new statistical foundation, targeting Asia and the Middle East, we have decided to assemble and maintain a list of the 500 most powerful computer systems in this region. Similar to the Topp500 list, our list is being compiled twice a year during GITEX exhibition in Riyadh and Dubai, with the help of high-performance computer experts, computational scientists, manufacturers, and the Internet community in general. We have also used parts of statistical lists published by others, such as the global Top500 list.

With the advancement of high-performance computers and large computing machines, manufacturers, users, and potential users wish to know not only the number of systems installed, but also the location of the various supercomputers within the high-performance computing community, and the applications for which a computer system is being used. Such statistics can facilitate the establishment of collaborations, the exchange of data and software, and provide a better understanding of the high-performance computing technology in the region.

The Asian top 500 list is compiled twice a year during Gitex exhibition in Riyadh and Dubai, with the help of high-performance computer experts, computational scientists, manufacturers, and the Internet community in general.

In the upcoming list, which is called the TOP500 Asia, supercomputers will be ranked by their performance using the LINPACK Benchmark, which is the de facto standard tool to measure the supercomputer's computational power in the high performance computing industry. The list maintainers intend to continue to update this list half-yearly and, in this way, to keep track with the evolution of computers in Asia continent.