I wouldn't say that cuda is more "optimized" but last I checked it does have better support for pointers, recursion, spawning new threads from inside of existin

I wouldn't say that cuda is more "optimized" but last I checked it does have better support for pointers, recursion, spawning new threads from inside of existing threads etc. This allows for more sophisticated data structures and algorithms that are impossible to replicate exactly in openCL. A Barnes-Hut simulation [1] for example is way tougher to build in OpenCL. [1] http://en.wikipedia.org/wiki/Barnes%E2%80%93Hut_simulation http://en.wikipedia.org/wiki/Barnes%E2%80%93Hut_simulation