We're generating a lot of data using our PyCuda simulations, and we've gotten to the point where analysis of the simulation output takes far more time than generating said output. I was thinking of collecting PyCuda accelerated statistics / spectral analysis functions into one place for convenience. Hopefully most of these would just be a few lines long and define operations in terms of the existing PyCuda base. I was just wondering if anything like this already existed ?
Sincerely, mrule. p.s.: Does PyCuda have access to the Cuda FFT implementation ? I think this was answered before on list, and the answer is no. Is there any GPU accelerated FFT that is easily accessible from PyCuda ?
_______________________________________________ PyCUDA mailing list [email protected] http://tiker.net/mailman/listinfo/pycuda_tiker.net
