I found in file adv_batch_mod.F90 that there are so many loops in the function adv_batch_accum_wind. Under the guidance of AI assistant, I tried to add OpenMP for them. However, the results are 9s slower than before on average.
Maybe I just need to calm down to learn some knowledge of multi-threading.
I found in file adv_batch_mod.F90 that there are so many loops in the function adv_batch_accum_wind. Under the guidance of AI assistant, I tried to add OpenMP for them. However, the results are 9s slower than before on average.
Maybe I just need to calm down to learn some knowledge of multi-threading.