GPU Computing is a proven technology providing both impressive speed up and energy efficient parallel computing power, from desktop workstations routinely used by engineers and researchers to very large hybrid clusters available in supercomputing centers.
HMPP is the leading software product enabling the rapid adoption and deployment of GPU empowered systems.  Porting and optimizing an application on GPU is easily done through the insertion of directives (Fortran code) or pragmas (C code) into the source code. The HMPP framework and code generator then automatically generates an efficient target code. Hybrid applications built with HMPP easily scale to multi-GPU systems by detecting at runtime the presence and availability of GPUs before offloading computations into them.
Based on its long standing expertise in parallel computing, CAPS also provides training and consulting services enabling customers to smoothly introduce GPU Computing within their organization.

CAPS and BULL team up to provide a comprehensive solution enabling a fast time-to-production and minimal porting effort for customers of bullx product line including ®Xeon®-based servers and NVIDIA® Tesla™ GPUs.

To speed-up GPU Computing adoption, BULL, NVIDIA and CAPS have launched the CAPS Compute Lab initiative (see CAPS Compute Lab for more details). CAPS Compute Lab will help any organization seeking to introduce NVIDIA® Tesla™  accelerated solutions to gain hands on experience using a Tesla empowered BULL cluster with training and support by CAPS experts. The CAPS Compute Lab package enables users to assess the speed up that they could achieve using a Tesla accelerated cluster and to document a ROI analysis for a future acquisition of such systems.

“We are pleased to have CAPS as partner supporting BULL HPC product line including NVIDIA Tesla GPUs, providing with its software product HMPP an efficient solution for the porting and optimization of legacy applications in C & Fortran on our machine. BULL and CAPS are joining their technologies and expertise to provide common customers and ISVs with an end-to-end solution enabling the deployment of GPU accelerated BULL solutions."

Fabio Gallo, VP of HPC Solutions at BULL