Microsoft HPC Pack
What’s Microsoft HPC Pack
Microsoft HPC Pack brings the power of high-performance computing (HPC) to the commercial mainstream. The centralized management and deployment interface helps to simplify deployment for both large and small compute clusters and provide a simple and effective management experience in order to increase cluster administrator productivity.
HPC Pack includes a highly scalable job scheduler that provides support for interactive Service-Oriented Architecture (SOA) applications using High Performance Computing for Windows Communication Foundation (HPC for WCF) and parallel jobs using the Microsoft Message Passing Interface (MS-MPI). Essential applications from key independent software providers (ISVs) can be run on the cluster to help you meet your business needs in a timely, cost-effective, and highly productive manner.
Integration with the Windows operating system helps to provide seamless security, storage, and desktop access to cluster resources and management. Microsoft Visual Studio provides parallel debugging capabilities for use with HPC Pack, and MS-MPI is now integrated with the Event Tracing for Windows infrastructure. This consolidates application, networking, and operating system events from many compute nodes into a single, time-correlated record to speed debugging.
HPC Pack includes an advanced job scheduler, rapid on-premises node deployment options using Windows Deployment Services, the ability to scale cluster capacity with nondedicated on-premises computers or with Windows Azure nodes (beginning in HPC Pack 2008 R2 with Service Pack 1), and a management interface built on the System Center user interface (UI), which has support for Windows PowerShell as a preferred scripting language. In HPC Pack, the head node takes advantage of Windows Server failover services as well as the failover clustering capabilities of Microsoft SQL Server for cluster failover and redundancy.