telecom • networking • design

an EventHelix.com blog


Leave a comment

GPU Based Parallel Computing with C++ AMP

C++ AMP lets you write C++ code that can directly target the GPU on the machine. Imagine being able to mix application code and signal processing code in a single program. The C++ AMP compiler takes care:

  1. Compiling the GPU code and loading it on the GPU
  2. Passing the data to from the application to the GPU
  3. Triggering the GPU computation
  4. Getting the results from the GPU and passing them back to the application.
The GPU code could perform performance intensive signal processing like FFT, turbo decoding and matrix multiplication.

http://channel9.msdn.com/posts/AFDS-Keynote-Herb-Sutter-Heterogeneous-Computing-and-C-AMP/player?w=512&h=288