3 ms·
I'd like to know more about how they are integating GPUs. I've been looking at CODA (nVidia library to use the GPU) and it looks mighty interesting.
by urlwolf 18y ago
I'd like to know more about how they are integating GPUs. I've been looking at CODA (nVidia library to use the GPU) and it looks mighty interesting.
- ComputerGuru 18y agoTake a look at the link I posted below - OpenCL can be accomplished by using a thread scheduler that's "aware" of asymmetric processing capabilities across the different cores (and treating the GPU as a usable processing core in non-OpenGL code). I suspect they'll be using ULE or something similar to it to pull this off.