Hi Jim -
You can time Copperhead programs the same way you time Python
programs, with the time module.
Or, if you're interested in just the CUDA kernel execution time, you
can use the CUDA profiler like with other CUDA programs.
I'm hoping to push an update next week to the main repository that
drastically reduces runtime overhead from the Python side, the current
main Copperhead branch has a lot of inefficiencies in the runtime that
have been removed - I just have to eliminate the last few bugs before
making the update public.
Let me know if you run into any problems, the Copperhead runtime and
compiler is still very buggy - but I'd be happy to help out if I can.
- bryan