Servers

CUDA: Speeding Up Parallel Computing

Download Now Free registration required

Executive Summary

CUDA is a computing engine in NVIDIA GPUs that is available to programmers using common programming languages. It's intended for developing parallel processing applications. The CUDA parallel-computing architecture supports many computational interfaces such as OpenGL and it is being used to speed up a large variety of applications in cryptography, computational biology and many other fields. The arrival of the multi-core CPUs many-core GPUs meant that the world as people know it was about to change. The processor chips that people know about have been transformed to parallel systems. Graphic cards are now not only advanced enough to function as good hardware accelerators: they also perform in a superior way than all the other alternatives that are currently being offered on commodity hardware platforms.

  • Format: PDF
  • Size: 21.4 KB