search

jadot-bp / fastsum_gpu

Repository for the "FAST" FASTSUM Analysis Software Tools prepared during the 2024 DiRAC GPU Workshop.
1 stars 0 forks source link

Non-Parallel Timers #7

Open RJaBi opened 8 months ago

RJaBi commented 8 months ago

The timer gives the core*real time, i.e. the amount of processor time. This is not really what we want.

This is particularly evident in the openMP branches. I am unsure about the parallel CPP branch. Unfortunately, this might be a thing that needs to be fixed across every different paradigm?

RJaBi commented 8 months ago

Have added this for the wrt.cpp file which measures Rx1 wilson loops. Have done this in both the feature_parallel_cpp and feature_openmp branches.

Fixed in commit ee31bd182fadfba6537b07dd05bf0c76057a478c for the feature_parallel_cpp branch Fixed in commit d9ed6a5c5c65820fcb6c1d1ad8ff61f72ec27f9d for the feature_openmp branch

The method in feature_parallel_cpp is probably robust (and usuable for i.e. the SYCL as well), though hasn't actually been tested for code running in parallel cause that's difficult on cosma with the standard parallel paradigm.