EXPERIMENTAL RESULTS ABOUT MPI COLLECTIVE COMMUNICATION OPERATIONS
Abstract
Collective communication performance is critical in a number of MPI applications. In this paper we focus on two widely used primitives, broadcast and reduce, and present experimental results obtained on a cluster of PC connected by InfiniBand. We integrated our algorithms in the MPICH library and we used MPICH implementation of broadcast and reduce primitives to compare the performance of our algorithms based on α-trees. Our tests show that the MPICH implementation can be improved.
References
- The International Journal of Supercomputer Applications and High Performance Computing 8(3/4), (1994). Google Scholar
- Concurrency: Practice and Experience 10(5), 359 (1998). Crossref, ISI, Google Scholar
- M. Bernaschi, G. Iannello, M. Lauria, Experimental results about MPI Collective Communication Operations Proceedings of HPCN99, P. Slot, M. Bubak, A. Hoekstra, and B. Hertzberger editors Lecture Notes in Computer Science (Springer) n. 1593 . Google Scholar
- A. Bar-Noy and S. Knipis, Designing Broadcasting Algorithms in the Postal Model for Message-Passing Systems, Procs. of the 4th Annual ACM Symp. on Parallel Algorithms and Architectures, June 1992 11–22 . Google Scholar
- R. M. Karp et al., Optimal Broadcast and Summation in the LogP Model, Procs. of the 5th Annual ACM Symp. on Parallel Algorithms and Architectures, June 1993, pp. 142–153 . Google Scholar
- , Recent Advances in Parallel Virtual Machine and Message Passing Interface,
Lecture Notes in Computer Science 2840 , eds.Jack Dongarra , Domenico Laforenza and Salvatore Orlando (Springer Verlag, Venice, Italy, 2003) pp. 257–267. Crossref, Google Scholar - M. Shroff, R. A. van de Geijn, CollMark: MPI Collective Communication Benchmark, tech. rep., The University of Texas at Austin, Department of Computer Sciences, December 1999 . Google Scholar
- H. Marcel, On Optimizing collective Communication, tech. rep., The University of Texas at Austin, Department of Computer Sciences, May 2003 . Google Scholar
- http://www.infinibandta.org/specs . Google Scholar
- Int'l Journal of Parallel Programming 32, 2 (2004). Google Scholar
- S. S. Vadhiyar, G. E. Fagg, J. Dongarra, Automatically tuned collective communications, Proceedings of the 2000 ACM/IEEE conference on Supercomputing . Google Scholar
- , Recent Advances in Parallel Virtual Machine and Message Passing Interface,
Lecture Notes in Computer Science 1697 , eds.J. Dongarra (Springer Verlag, Barcelona, Spain, 1999) pp. 469–476. Crossref, Google Scholar - R. Rabenseifner, New optimized MPI reduce algorithm, http://www.hlrs.de/organization/par/services/models/mpi/myreduce.html . Google Scholar
- Journal of Systems Architecture 49, 3 (2003). ISI, Google Scholar
- , Recent Advances in Parallel Virtual Machine and Message Passing Interface,
Lecture Notes in Computer Science 2474 , eds.J. Dongarra (Springer Verlag, 2002) pp. 392–400. Google Scholar - S. Sistare, R. vandeVaart, and E. Loh, Optimization of MPI collectives on clusters of large-scale SMPs. in Procs. of SC99: High Performance Networking and Computing, November 1999 . Google Scholar
- E. W. Chan, M. F. Heimlich, A. Purakayastha, and R. A. van de Geijn, On Optimizing Collective Communication, In Proc. of the 2004 IEEE International Conference on Cluster Computing (Cluster 2004), September 2004 . Google Scholar
- International Journal of High Performance Computing Applications 1(19), 49 (2005). Google Scholar


