MPIPerformance

From BCCD 3.0

• OpenMPI 1.2.6, gromacs 3.3.3, fftw 3.1.2 on three cairo nodes, one process per node:
  • real 27m3.916s, user 0m0.871s, sys 0m0.065s
• MPICH2 1.0.7, gromacs 3.3.3, fftw 3.1.2 on three cairo nodes, one process per node:
  • real 26m58.225s, user 0m0.250s, sys 0m0.042s