DocumentCode :
1647884
Title :
Investigating the performance of two programming models for clusters of SMP PCs
Author :
Cappello, Franck ; Richard, Olivier ; Etiemble, Daniel
Author_Institution :
Lab. de Recherche en Inf., Univ. de Paris-Sud, Orsay, France
fYear :
2000
fDate :
6/22/1905 12:00:00 AM
Firstpage :
349
Lastpage :
359
Abstract :
Multiprocessors and high performance networks allow building CLUsters of MultiProcessors (CLUMPs). One distinctive feature over traditional parallel computers is their hybrid memory model (message passing between the nodes and shared memory inside the nodes). We evaluate the performance of a cluster of dual processor PCs connected by a Myrinet network for NAS benchmarks using two programming models: a Single Memory Model based on the MPICH-PM/CLUMP library of the RWCP and a Hybrid Memory Model using MPICH-PM and OpenMP. MPI programs are used as the reference in all experiments involving programming models. We compare dual processor node configurations speedup versus uniprocessor node configurations for each model. We demonstrate that the superiority of one model over the other depends on the features of the applications. In particular, we detail the speedup results from breakdowns of the benchmark execution times and from measurements of hardware counters
Keywords :
message passing; multiprocessing systems; performance evaluation; workstation clusters; Myrinet network; NAS benchmarks; clusters of SMP PCs; dual processor node configurations speedup; hardware counters; high performance networks; hybrid memory model; message passing; multiprocessors; performance; programming models; Costs; Counting circuits; Hardware; Hidden Markov models; High-speed networks; Message passing; Parallel machines; Particle measurements; Personal communication networks; Read only memory;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
High-Performance Computer Architecture, 2000. HPCA-6. Proceedings. Sixth International Symposium on
Conference_Location :
Touluse
Print_ISBN :
0-7695-0550-3
Type :
conf
DOI :
10.1109/HPCA.2000.824364
Filename :
824364
Link To Document :
بازگشت