Windows HPC Server 2008, the successor to Windows Compute Cluster Server 2003, is based on the Windows Server 2008 operating system and is designed to increase productivity, scalability and manageability. Windows HPC Server 2008 has been renamed to reflect its readiness to tackle the most challenging HPC workloads. Key features are new high-speed networking, highly efficient and scalable cluster management tools, advanced failover capabilities, a service oriented architecture (SOA) job scheduler, and support for partners’ clustered file systems. The beta is now available for download at http://www.microsoft.com/hpc; the final version will be generally available in the second half of 2008.
“With the new advancements, Windows HPC Server 2008 can allow customers to achieve the levels of scalability and performance of the most efficient clusters in the Top500 benchmark while making it dramatically more productive to deploy, utilize and integrate the advanced HPC clusters within their environment,” said Kyril Faenov, general manager of HPC at Microsoft. “By upgrading to Windows HPC Server 2008 on our 2,048-core production test cluster, we increased the LINPACK performance by 30 percent and were able to deploy and validate the cluster in less than two hours using out-of-the-box software. Expanding beyond traditional MPI-based HPC applications, Windows HPC Server 2008 enables support for high-throughput SOA applications with its advanced Web service routing capability and paves the way for bringing HPC capabilities to a broad range of enterprise applications.”
This efficiency is currently being demonstrated at the Holland Computing Center in the Peter Kiewit Institute at the University of Nebraska, one of the largest clusters in the world. The new 1,151-node Windows-based cluster is expected to enhance the curriculum and resource capabilities by providing compute cycles for a broad variety of government, research and industry uses.
Press Release and download beta here.