I have not had any success using OMP_PROC_BIND or GOMP_CPU_AFFINITY on Windows 7 or 10 OS.
Has anyone been able to demonstrate this can improve performance ?
For my type of shared memory/data calculation, involving a large shared array, OpenMP as a shared memory model looks to have an advantage over (my understanding of) MPI.