How to profile (visually) a CUDA code which is implemented in a python package through C extension?

Asked Jan 07 '13 at 23:09

Active Jan 07 '13 at 23:09

Viewed 46 times

Possible Duplicate:
How to profile PyCuda code with the Visual Profiler?

The CUDA visual profiler (nvvp) requires an executable entry for profiling, but my CUDA code is implemented in a python package by through C extension. Is there anyway to do the CUDA profiling in this python package? Thanks!

edited May 23 '17 at 10:24

Community

asked Jan 07 '13 at 23:09

Hailiang Zhang

17,604
23
71
117

Is there a reason that you cannot specify to NVVP the application that is loading the python C extension as the application to profiler? – Greg Smith Jan 07 '13 at 23:29
@Greg Smith Thanks for your reply, but NVVP can only be initiated with an executable, and my application entrance is a python script and has to be invoked by python. May I missed something here... – Hailiang Zhang Jan 08 '13 at 00:15
How about launching with python.exe? – Roger Dahl Jan 08 '13 at 00:17
Yes, it is a duplicate post. Please close it. – Hailiang Zhang Jan 08 '13 at 19:17

How to profile (visually) a CUDA code which is implemented in a python package through C extension?

0 Answers0