is there anybody has implementation of LINPACK in pycuda??
how to count gpu flops compared to CPU like in NVIDIA slides and book??
_______________________________________________ PyCUDA mailing list [email protected] http://lists.tiker.net/listinfo/pycuda
