(NASDAQ: VOLT), a leading provider of scale-out data center fabrics, today announced the availability of Voltaire’s Fabric Collective Accelerator™ (FCA™) solution for environments using Platform MPI. Voltaire FCA is a software and hardware solution that speeds up message passing interface (MPI) performance over x86 based clusters to accelerate performance of high performance computing (HPC) applications such as reservoir modeling, fluid dynamics, crash analysis and others.
In a new benchmark test conducted by HP, Voltaire FCA software improved performance of ANSYS FLUENT between 14 and 30% and CD-Adapco Star-CD AClass by 10%. The 192 core benchmark environment consisted of 96 HP BL460 blade servers, Voltaire Grid Director™ 40 Gb/s InfiniBand switches and Platform MPI.
Voltaire FCA software accelerates MPI collective operations by using Voltaire switches and their on-board processors to offload significant parts of group communication onto the switching fabric while Voltaire Unified Fabric Manager™ software (UFM™) orchestrates an efficient, topology-based collective flow. Working in concert, these products ensure that all bottlenecks are removed, at both the server and interconnect levels. The computational acceleration is achieved transparently without requiring changes to the application.
Platform MPI is a high performance, production–quality implementation of the Message Passing Interface (MPI). It is widely used in the high performance computing industry and is considered the de facto standard for developing scalable, parallel applications. Platform MPI supports the broadest range of industry standard platforms, interconnects and operating systems helping ensure that parallel applications can run anywhere.
“Gaining a ten to thirty percent improvement in application performance is very significant and can mean substantial gains in competitive advantage for a business,” said Asaf Somekh, Vice President of Marketing, Voltaire. “The integration of Voltaire FCA software and Platform MPI enables a broader range of HPC customers to leverage these new performance advancements.”
“Scaling parallel application performance over large HPC clusters is a known challenge due to the inherent limitations of parallel communication,” said Tripp Purvis, Vice President, Business Development, Platform Computing. “The combination of Platform MPI and Voltaire FCA enables a quantum leap in MPI-and HPC- application performance over x86-based clusters.”