On Wed, 12 Oct 2011 00:00:39 +1300, Igor <rychphd(a)gmail.com> wrote:
Is it possible to compile some thrust code for example from here
...or will PyCUDA be always limited to compiling kernels only? If
that's the case, will PyCUDA reimplement thrust's code such as
thrust/extrema.h? I am particularly interested in finding max/min
elements in a GPU vector. Thrust allows it, and PyCUDA doesn't seem
For now, PyCUDA is kernel-only.
How about this here?