openmp / sources Goto Github PK

Are there any known issues on scalability in terms of number of
cores/threads for OpenMP?

Let's say if we have 96 threads on a single server - would this scale up?
We're using some R libraries that use OpenMP
and with higher thread counts see some delays and generally
unexpected issues with scaling things up.

I have been writing a code and have tried to parallelize it using #pragma omp parallel for. Everything works fine till about 100000 iterations but whe i increase the number of iterations to about 200000, it compiles without errors but while running shows the error "segmentation fault(core dumped) " immidiately.
I have been told that it might be local/global variable issue of the iteration variables but i think it is unlikely because it runs just fine if the number of iteration is lesser.
Also i have been told that it might be stack issue which i have no idea about