STORAGE

Stanford University has been awarded with Storage Networking World's "Best Practices in Storage" Award for their successful design, implementation and management of a high performance computing (HPC) infrastructure using Panasas ActiveStor Parallel Storage Clusters. The award was determined by IDG's Computerworld Magazine's advisory group and alumni panel. Stanford received this recognition for its deployment of Panasas' parallel clustered storage solutions in their world class research facility which focuses on solving engineering challenges that require massively parallel computing resources for large-scale simulations such as computational fluid dynamics (CFD). Panasas' solution has allowed Stanford scientists to perform quick, efficient analysis of the physics included in simulations allowing them to make critical decisions on prioritizing future calculations and simulations. "Panasas' ActiveStor Parallel Storage Cluster has helped Stanford's HPC center win this prestigious award by dramatically improving our ability to tackle complex simulations and speed results that were not possible with previous storage solutions," said Steve Jones, HPC manager at Stanford University. "Stanford's HPC center was able to integrate Panasas parallel storage into the Center's infrastructure in less than two hours, with minimal administrative requirements. We now receive accelerated results across a variety of complex scientific applications. With Panasas, large-scale storage clusters can be assembled and running within days instead of weeks or months." All Panasas storage clusters contain the object-based PanFS parallel file system. Using Panasas storage, customers can achieve maximum application performance while reducing total cost of ownership by driving the cost and complexity out of their storage infrastructure. The fully-parallel Panasas DirectFlow protocol provides direct access between Linux cluster nodes and Panasas storage to eliminate costly performance bottlenecks that legacy storage systems can create when used with parallel applications. Additionally, a single global namespace dramatically simplifies cluster management. "Recognizing Stanford for their HPC Center implementation illustrates how Panasas parallel storage solutions provide the highest performance to tackle demanding simulation and modeling applications, which results in substantially faster research and development," said Len Rosenthal, chief marketing officer at Panasas. "We are honored to play a significant role in Stanford's preeminent High Performance Computing Center and look forward to a continued close partnership with Stanford."