ARCHIVE

50% Performance Improvements for Electromagnetic Simulations with NVIDIA’s CUDA Language Powering Acceleware’s Next Generation of Software Acceleration Products: 

Acceleware today announced a new CUDA (Compute Unified Device Architecture) based Acceleware solution for accelerating EM simulations. This new software library has delivered application performance gains of up to 50% compared to the current OpenGL version.  This next generation version is compatible with both 8- and 10-series NVIDIA GPUs, delivering significant speed-ups to customers without having to upgrade hardware. This functionality is offered free of charge to users with active Acceleware maintenance.

“CUDA allows us to better exploit the full potential of NVIDIA’s Tesla GPUs, and our performance proves it. There is no question that this is the right tool for developing high performance scientific libraries and applications,” said Robert Miller, Director of Product Management at Acceleware.

This new Acceleware software uses NVIDIA’s CUDA programming language that was designed to solve complex computational problems, enabling users to access multi-core parallel processing technology. Users will benefit from the increased acceleration performance and the ability to run 50% larger simulation problems on 10-series NVIDIA hardware thanks to better memory access. CUDA also provides complete access to the NVIDIA Tesla GPU line, providing a compelling price to performance value for end users.

“Acceleware’s new CUDA libraries provide our SEMCAD X customers with even better performance from their existing GPU hardware. Furthermore, it empowers amazing and unrivalled performance in both computation size and runtime for the latest series of NVIDIA GPUs. These results further greatly justify our investment in GPU computing and our partnership with Acceleware - providing leading-edge technology that has become absolutely indispensable to the whole CEM community,” said Nik Chavannes, Director of Software at SPEAG.

Acceleware’s CUDA-based solvers take advantage of highly parallel NVIDIA compute hardware to offer performance gains while performing lengthy EM design simulations. End users will continue to benefit from faster design cycles and more efficient use of engineering resources.