CUDA @ Scalable Graphics

What is CUDA?

NVIDIA CUDA™ technology enables you to exploit the computation power of the Graphics Processing Units (GPUs). GPUs possess a highly parallel architecture, making them extremely more effective than general-purpose CPUs for a wide range of complex algorithms. Through its extension for the C-language and the high performance computing scientific libraries, CUDA offers an efficient tool to harness the computational power of GPUs.

Visit the official CUDA page for a more detailed description about this emerging technology.

CUDA Training and Expertise

Getting the maximum performance using CUDA might be a daunting task. First, it requires some parallel computing background, and a good understanding of the hardware architecture. Second, one need to master the do's and do'nt of CUDA.

Scalable Graphics has more than ten years of experience in parallel computing and GPU programming. Whether you need a little help to get you started, or are looking for sharp technical expertise, we have the answers. Check our training page for more information on the courses we offer.

We also carry CUDA specific developments for our customers. In particular, we have a long term proficiency in parallel algorithms for data compression, simulation (finite elements and stochastic methods), visualization and large datasets. We help you define tailor-made hardware configurations matching your problems size and performance needs. Our expertise ranges from single systems to NVIDIA Tesla based PC clusters.