Has anyone here used double precision scientific computing with new generation (e.g. K20) GPUs through Python?
I know that this technology is rapidly evolving, but what is the best way to do this currently? GPU is out of scope for the popular scientific Python libraries numpy and scipy, and I had wanted to use theano but it seems to use only float32 precision for GPU. I am aware that google can provide search results for python gpu, but I am hoping for more insight than a list of projects which may or may not be on their way to meet their maker.