OpenCL Downloads

Nvidia has an OpenCL beta download available. Nvidia's OpenCL driver will only allow you to run on Nvidia devices. You must register as a Nvidia developer to gain access. For more information

http://www.nvidia.com/object/cuda_opencl.html

AMD has an OpenCL beta download available. AMD's OpenCL driver will only allow you to run on AMD CPUs. They will be adding support for running on ATI GPUs towards the end of the year. You must register as an AMD developer to gain access. For more information

http://developer.amd.com/gpu/ATIStreamSDKBetaProgram/Pages/default.aspx

OpenCL Tutorials

AMD has put together a very basic OpenCL tutorial using the proposed C++ bindings to OpenCL. This tutorial can be found at

http://developer.amd.com/gpu/ATIStreamSDK/pages/TutorialOpenCL.aspx

AMD is working with Mindshare to develope a 3 day class on OpenCL programming. The class is not yet available but should be soon. For more information

http://www.mindshare.com/learn/?section=0A6B150A00B5

Nvidia webinars that provide an interactive learning experience. This series will cover the basics of data parallel computing on GPU’s leveraging NVIDIA’s CUDA architecture . Tutorials will cover many topics including C for CUDA, programming to the OpenCL™ API , using Direct X Compute and performance optimization techniques, presented by NVIDIA Developer Technology Engineering team and NVIDIA staff online to answer Questions.

You can register for these at http://developer.nvidia.com/object/gpu_computing_online.html

OpenCL Documentation

The OpenCL specification is maintained by the Khronos group. It can be found at

http://www.khronos.org/registry/cl/specs/opencl-1.0.43.pdf

The SIGGRAPH ASIA Khronos OpenCL presentation by Neil Trevett of NVIDIA can be found at

http://www.khronos.org/developers/library/overview/opencl_overview.pdf

There is a quick reference card for OpenCL 1.0 API that can be found at

http://www.khronos.org/files/opencl-quick-reference-card.pdf

This guide (The Nvidia OpenCL JumpStart Guide) will help you to start developing GPU accelerated applications today, using C for CUDA compute kernels and the CUDA Driver API in ways that that will make it easy to transition to OpenCL when you are ready.

The Nvidia OpenCL JumpStart Guide can be found at http://developer.download.nvidia.com/OpenCL/NVIDIA_OpenCL_JumpStart_Guide.pdf

AMD has put together an introduction to OpenCL that briefly goes over the anatomy of OpenCL 1.0 in its three main parts and shows a simple code example. It is not meant to be a detailed tutorial for OpenCL 1.0.

http://sites.amd.com/us/fusion/apu/Pages/fusion.aspx

CUDA Education

CUDA is currently being taught at over 50 universities worldwide. Some of the more notable are Harvard, MIT, University of Illinois, and the Chinese Academy of Science Institute for Computing Technology. You can find a complete list of institution and course offerings at

http://www.nvidia.com/object/cuda_university_courses.html

CUDA Zone

Nvidia has put together a webportal for CUDA developers called the CUDA Zone. This site contains documentation, software, examples of what others are doing with CUDA, tutorials, and the official CUDA forums.

http://www.nvidia.com/object/cuda_home.html#

CUDA Downloads

The CUDA driver, toolkit, SDK, and code examples for all supported platforms can be found at

http://www.nvidia.com/object/cuda_get.html

Cuda Documentation

Tom Halfhill has put together a very good overview of CUDA.

http://www.nvidia.com/docs/IO/55972/220401_Reprint.pdf

The official Nvidia documation for CUDA including a QuickStart Guide, Programming Guide, and Reference Manual for all versions and supported platforms can be found at

http://www.nvidia.com/object/cuda_develop.html

The Nvidia CUDA C Programming Best Practices Guide can be found at

http://developer.download.nvidia.com/compute/cuda/2_3/toolkit/docs/NVIDIA_CUDA_BestPracticesGuide_2.3.pdf