Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Starfyre

macrumors 68030
Original poster
Nov 7, 2010
2,905
1,136
Isn't Nvidia's CUDA meant to do computations? What is the benefit of OpenCL besides the fact that it is part of the iGPU? When would a "Pro" use the 750M vs the Iris Pro for each of those?
 
OpenCL and Cuda are programming APIs. A developer who wants to write code that is run on the GPU has to decide whether he wants to use OpenCL or Cuda. Nvidia started cuda but nobody every jumped on board. I am not even sure if there is anything on OSX that supports Cuda. OpenCL is platform independent and is supported by all GPU manufacturers (Intel, Nvidia, AMD, ..). OpenCL is an open standard and probably the only API developers will focus for that kind of functionality.

They are basically programming languages with Cuda only being supported by Nvidia GPUs and nobody else.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.