Brook+, OpenCL and CUDA

Discussion created by packoman on Dec 10, 2008
Capability, Compatibility and the future -> a question


as I am currently planning to get into GPGPU computing (still need to buy a compliant graphics card) and with the Specs of OpenCL now released, there is a question that came to my mind.

I read the following on the engadget site about CUDA and Nvidias view of OpenCL:

"NVIDIA feels that the basic ideas behind CUDA are reinforced by OpenCL -- the basic model is the same, although OpenCL is a lower-level architecture that requires coders to do their own memory management, etc., while CUDA allows non-professional programmers like scientists and researchers to program GPGPU code.

In fact, in order to meet the needs of the 25,000 active CUDA devs, NVIDIA is bumping up the CUDA release schedule, with three releases scheduled for 2009, culminating in CUDA 3.0."

So how is the stance on the side of AMD/ATI? I read somewhere that AMD will ease the porting of apps  from Brook+ to OpenCL for developers. Does that mean AMD will depriciate the Brook+ framework completely or will they also take a similiar approach like NVIDIA?

I hope understand the quote correctly in that they are planning, that CUDA at some point will use OpenCL as a "back-end" so that CUDA applications in principle could also be compiled for ATI-cards.

If so will this also be the case for Brook+? Or will the only possibilty be to code directly in OpenCL?

Up to know I don't really know alot about these things, so this question might sound stupid. Hopefully somebody can enlighten me.