Nvidia releases a GPU accelerated engineering simulation application

CHIP DESGINER Nvidia has released a GPU accelerated version of the MSC Nastran 2012 Finite Element Analysis (FEA) application.

The application is used in a number of engineering simulation tasks and is available for both Windows and Linux. The firm touts a performance improvement of up to five times when compared to the previous version.

"GPU-enabled performance improvements have the potential to transform engineering analysis and design-optimization procedures," said Dr Ted Wertheimer, senior director of product management at MSC Software. "CUDA-based GPU acceleration in MSC Nastran 2012 will speed up performance by 1.5-5x for a range of models and industries, helping customers improve their workflow and deliver better products to market, faster."

By adding Nvidia GPU acceleration users will benefit from more realistic models and higher quality simulations. This will allow breakthroughs for aircraft, spacecraft and other mission critical systems, according to Nvidia.

The firm also announced yesterday at the GPU Technology Conference in Beijing that it is releasing the source code for its low level virtual machine (LLVM) based CUDA compiler.

Andrew Cresci, general manager of strategic alliances at Nvidia said, "We are expanding our joint development with MSC to include GPU acceleration of complex math kernels, allowing us to address NVH and large dynamics problems for our customers." µ