Open MPI: A High Performance, Flexible Implementation of MPI Point-to-Point Communications
Abstract
Open MPI's point-to-point communications abstractions, described in this paper, handle several different communications scenarios, with a portable, high-performance design and implementation. These abstractions support two types of low-level communication protocols – general purpose point-to-point communications, like the OpenIB interface, and MPI-like interfaces, such as Myricom's MX library. Support for the first type of protocols makes use of all communications resources available to a given application run, with optional support for communications error recovery. The latter provides a interface layer, relying on the communications library to guarantee correct MPI message ordering and matching. This paper describes the three point-to-point communications protocols currently supported in the Open MPI implementation, supported with performance data. This includes comparisons with other MPI implementations using the OpenIB, MX, and GM communications libraries.
References
G. Burns , R. Daoud and J. Vaigl , LAM: An Open Cluster Environment for MPI, Proceedings of Supercomputing Symposium (1994) pp. 379–386. Google Scholar-
Graham E. Fagg , Fault tolerant communication library and applications for high perofrmance , Los Alamos Computer Science Institute Symposium ( 2003 ) . Google Scholar -
E. Garbriel , Open MPI: goals, concept, and design of a next generation MPI implementation , Proceedings, 11th European PVM/MPI Users' Group Meeting ( 2004 ) . Google Scholar A. Geist , MPI-2: Extending the Message-Passing Interface, Euro-Par '96 Parallel Processing (Springer Verlag, 1996) pp. 128–135. Google Scholar- International Journal of Parallel Programming 31(4), (2003), DOI: 10.1023/A:1024504726988. Google Scholar
- Parallel Computing 22(6), 789 (1996), DOI: 10.1016/0167-8191(96)00024-5. Crossref, ISI, Google Scholar
- Journal of Grid Computing 1, 133 (2003), DOI: 10.1023/B:GRID.0000024071.12177.91. Crossref, ISI, Google Scholar
- , MPI: A Message Passing Interface, Proc. of Supercomputing '93 (IEEE Computer Society Press, 1993) pp. 878–883. Google Scholar
- Mpich2, argonne, http://www-unix.mcs.anl.gov/mpi/mpich2/ . Google Scholar
-
S. Pakin and A. Pant , VMI 2.0: A dynamically reconfigurable messaging layer for availability, usability, and management , Proceedings of The 8th International Symposium on High Performance Computer Architecture (HPCA-8) ( 2002 ) . Google Scholar - Quadrics, llc web page, http://www.quadrics.com/ . Google Scholar
-
G. M. Shipman , Infini-band scalability in open mpi , Proceedings, 20th IEEE International Parallel & Distributed Processing Symposium ( 2006 ) . Google Scholar Jeffrey M. Squyres and Andrew Lumsdaine , The component architecture of open MPI: Enabling third-party collective algorithms, Proceedings, 18th ACM International Conference on Supercomputing, Workshop on Component Models and Systems for Grid Applications, eds.Vladimir Getov and Thilo Kielmann (Springer, 2004) pp. 167–185. Google Scholar- Sun, llc web page, http://www.sun.com/ . Google Scholar
-
T. S. Woodall , Open MPI's TEG point-to-point communications methodology : Comparison to existing implementations , Proceedings, 11th European PVM/MPI Users' Group Meeting ( 2004 ) . Google Scholar -
T. S. Woodall , TEG: A high-performance, scalable, multi-network point-to-point communications methodolog , Proceedings, 11th European PVM/MPI Users' Group Meeting ( 2004 ) . Google Scholar


