GPU programming in a glance

On my MacBook I have a nVIDIA GeForce 8600M GT which is CUDA enabled, something I never bothered checking until very recently. nVIDIA provides online the required driver, SDK and additional “CUDA Developer”, as they call it, resources with lots of sample files to test the hardware of your system as well as actual code, including some parallel samples.

The CUDA toolkit seem to provide all you need to start with:

  • C/C++ compiler
  • Visual Profiler
  • GPU-accelerated BLAS library
  • GPU-accelerated FFT library
  • GPU-accelerated Sparse Matrix library
  • GPU-accelerated RNG library
  • Additional tools and documentation

It does also include OpenCL samples to play about. However, the OpenCL driver will need to be installed at first place. There’s a pre-release version and in order to download it you’d need to register yourself with nVIDIA. They have also published a book, “CUDA by example”, which is not for free apart from some fragments. Nevertheless, the sample codes of the book are free to download.

AMD / ATI have also their answer to CUDA, “ATI Stream“. From what I got it seems to support only OpenCL. I don’t have a Stream-supported ATI card at the moment so I couldn’t try that one.

To close this, there’s an interesting presentation that covers basics of GPUs and how to program them (CUDA based): Programming and optimization of applications for multiple GPU

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s