Performance of a Code Migration for the Simulation of Supersonic Ejector Flow to SMP, MIC, and GPU Using OpenMP, OpenMP+LEO, and OpenACC Directives

Listing 8

Iterations of the outer loops are joined to be executed in parallel and the internal operations are intended to be vectorized. As the inner loop exists, it is intended to be vectorized.

!$OMP DO SIMD COLLAPSE(2)

DO j=2,(nj-1)

  DO i=2,(ni-1)

    DO k=2,(nz-1)

    END DO

  END DO

END DO