MPI versus MPI+OpenMP on IBM SP for the NAS benchmarks

  • Authors:

  • Franck Cappello;Daniel Etiemble

  • Affiliations:

  • LRI, Université Paris-Sud, 91405, Orsay, France;LRI, Université Paris-Sud, 91405, Orsay, France

  • Venue:
  • Proceedings of the 2000 ACM/IEEE conference on Supercomputing
  • Year:
  • 2000

Quantified Score

Hi-index 0.00

Visualization

Abstract

The hybrid memory model of clusters of multiprocessors raises two issues: programming model and performance. Many parallel programs have been written by using the MPI standard. To evaluate the pertinence of hybrid models for existing MPI codes, we compare a unified model (MPI) and a hybrid one (OpenMP fine grain parallelization after profiling) for the NAS 2.3 benchmarks on two IBM SP systems. The superiority of one model depends on 1) the level of shared memory model parallelization, 2) the communication patterns and 3) the memory access patterns. The relative speeds of the main architecture components (CPU, memory, and network) are of tremendous importance for selecting one model. With the used hybrid model, our results show that a unified MPI approach is better for most of the benchmarks. The hybrid approach becomes better only when fast processors make the communication performance significant and the level of parallelization is sufficient.