Out of curiosity, are you (or anyone else here) familiar with how this Apple stuff actually works from a developer perspective - do you mess with actual OpenCL code, or do you use some Apple interfaces and don't benefit from knowing OpenCL itself?
(Asking because I'm likely going to get to know CUDA and/or OpenCL in the near future. I'd be using with generic C++, not something Apple specific.)
I know a cursory amount of C, OpenGL and OpenCL. Mostly all I need to touch are markup languages and JavaScript for Adobe scripting though.
CUDA might be a good idea but it's implementation in Adobe applications is so cruddy that I've been turned off it (also I don't really like its Nvidia only status but that's just the way of business.)
What're you going to r using it for.