Microsoft Windows HPC Server 2008 Clusters Scale at up to 85% Efficiency
Written on June 27, 2008 – 12:04 pm | by admin
Microsoft announces that a Windows HPC Server 2008 beta-based system now ranks among the top 25 supercomputers of the world. They have also announced a release-candidate version of the operating system software available for download towards the end of June. The final release version of this software will be available by the end of this year.
A Beta 2 version of the said software was used by NCSA (National Center for Supercomputing Applications) in a system which ranked 23 in the list of top 500 supercomputers. The NCSA achieved 68.5 TFLOPs and 77.7% efficiency on 9,472 cores, making this one of the most powerful supercomputing systems in the world. This also marks the first time a Microsoft operating system-based cluster made it into the top 25.
The computer scientists at Umea University, Sweden devised the other cluster using this software, achieved 46 TFLOP’s at an efficiency of 85% using 5,476 processing cores.
The operating system is based on a Windows Server 2008 foundation, but can scale to thousands of cores because of features like a new high-speed NetworkDirect RDMA, Microsoft’s new remote direct memory access interface, cluster management tools, a service-oriented architecture (SOA) job scheduler, and cluster interoperability through standards such as the High Performance Computing Basic Profile (HPCBP) specification produced by the Open Grid Forum (OGF).
For more information, please visit this page.
Tags: microsoft, server 2008, windows