R Systems Achieves Significant Speed-Up for Abaqus FEA from SIMULIA Using Flexible Cluster Configuration
Dassault Systèmes announced that R Systems, a technology-leading IT solutions provider, in cooperation with SIMULIA and Dana Holding Corporation, has completed a simulation benchmark using Abaqus FEA software and Windows HPC Server 2008.
“With Windows HPC Server 2008, Microsoft has made huge advancements in performance and scaling that give users of realistic simulation solutions more IT flexibility,” stated Brian Kucic, VP of business development for R Systems. “Using the analysis model provided by Dana and the same 32-node compute cluster, we evaluated the performance of Abaqus FEA software on both Linux and Windows. Switching between the two operating systems was straightforward and the performance of Windows HPC Server 2008 was highly competitive.”
Using 128 cores of a 256 core Intel Harpertown cluster with Quad Data Rate Infiniband, R Systems performed the benchmark study of a finite element analysis model provided by Dana to evaluate distributed memory performance of Abaqus FEA on Windows. The analysis, which normally takes a little over 2 days to complete using 8 cores, ran to completion in just under 3.5 hours.
“We were extremely pleased by the outcome of the R Systems study,” stated Frank Popielas, manager advanced engineering, Sealing Products Group, Dana Holding. “The combination of flexible computing clusters and the power of parallel processing enabled by SIMULIA and Microsoft will result in significant time and money savings for Dana and the manufacturing industry as a whole.”
“SIMULIA is focused on optimizing the performance of our robust and scalable simulation software for Windows HPC Server 2008,” stated Steve Crowley, director of product management, SIMULIA, Dassault Systèmes. “R Systems provided excellent computing resources for our team to assess the performance of Windows HPC Server 2008 using Dana’s model. We were highly impressed with the results and believe Windows HPC Server 2008 makes supercomputing more accessible by allowing our end-users to harness the power of distributed computing through a familiar Windows desktop environment.”
Running Abaqus FEA software on a Windows-based cluster provides engineers with an integrated, efficient, and easily scalable high-performance computing platform for solving some of the most complex simulation and multiphysics analysis equations in less time by leveraging the combination of Abaqus distributed-memory parallel technology and Windows clustering solutions.