Designing zero-copy message passing interface derived datatype communication over InfiniBand: Alternative approaches and performance evaluation

International Journal of High Performance Computing Applications

Volumn 19, Issue 2 SPEC. ISS., 2005, Pages 129-142

  Santhanaraman, Gopalakrishnan ^a   Wu, Jiesheng ^a   Huang, Wei ^a   Panda, Dhabaleswar K ^a  

^a Ohio State University   (United States)

Author keywords

Datatype; InfiniBand; MPI; Zero copy

Indexed keywords

BANDWIDTH; COMMUNICATION CHANNELS (INFORMATION THEORY); COMPUTER NETWORKS; DATA TRANSFER; INTERCONNECTION NETWORKS; NETWORK PROTOCOLS; SYSTEMS ANALYSIS;

DATATYPE; INFINIBAND CHANNEL SEMANTICS; MESSAGE PASSING INTERFACE (MPI); SEND GATHER RECEIVE SCATTER (SGRS); ZERO-COPY MESSAGE PASSING INTERFACE;

DATA COMMUNICATION SYSTEMS;

EID: 19644366328     PISSN: 10943420     EISSN: None     Source Type: Journal    
DOI: 10.1177/1094342005054259     Document Type: Article

References (14)

1. Ashworth, M. 1996. A report on further progress in the development of codes for the CS2. Deliverable D.4.1.b, F. Carbonnell, editor, GPMIMD2 ESPRIT Project, EU DGIII, Brussels.
2. Byna, S., Sun, X-H., Gropp, W., and Thakur, R. 2003. Improving the performance of MPI-derived datatypes by optimizing memory-access cost. Proceedings of the IEEE International Conference on Cluster Computing , Hong Kong, China, December.
3. Fryxell, B., Olson, K., Ricker, P., Timmes, F. X., Zingale, M., Lamb, D. Q., MacNeice, P., Rosner, R., and Tufo, H. 2000. FLASH: an adaptive mesh hydrodynamics code for modelling astrophysical thermonuclear flashes. Astro-physical Journal Supplement, 131:273.
4. Gropp, W., Lusk, E., Doss, N., and Skjellum, A. 1996. A high-performance, portable implementation of the MPI message passing interface standard. Parallel Computing, 22(6):789-828.
5. Gropp, W., Lusk, E., and Swider, D. 1999. Improving the performance of MPI-derived datatypes. Message Passing Interface Developer's and User's Conference (MPIDC), Atlanta, GA, March.
6. Liu, J., Wu, J., Kini, S. P., Wyckoff, P., and Panda, D. K. 2003. High performance RDMA-based MPI implementation over InfiniBand. Proceedings of the 17th Annual ACM International Conference on Supercomputing, San Francisco, CA, June.
7. Lu, Q., Wu, J., Panda, D. K., and Sadayappan, P. 2004. Employing MPI-derived datatypes to the NAS benchmarks: a case study. Technical Report OSU-CISRC-02/ 04-TR10, Department of Computer and Information Science, Ohio State University, February.
8. MPI Forum. 1994. MPI: a message-passing interface standard. International Journal of Supercomputer Applications and High Performance Computing, 8(3-4): 165-414.
9. Ross, R., Miller, N., and Gropp, W. 2003. Implementing fast and reusable datatype processing. Proceedings of the 10th European PVM/MPI Users' Group conference (EuroPVM/MPI), Venice, Italy, September 29-October 2.
10. Tipparaju, V., Santhanaraman, G., Nieplocha, J., and Panda, D. K. 2004. Host-assisted zero-copy remote memory access communication on InfiniBand. International Parallel and Distributed Processing Symposium (IPDPS'04), Santa Fe, NM, April.
11. Träff, J. L., Ritzdorf, H., and Hempel, R. 2000. The implementation of MPI-2 one-sided communication for the NEC SX. Proceedings of the International Conference on Supercomputing, Santa Fe, NM, May.
12. Worringen, J., Gaer, A., Reker, F., and Bemmerl, T. 2002. Exploiting transparent remote memory access for noncontiguous-and one-sided-communication. Workshop on Communication Architecture for Clusters 2002 (in conjunction with the International Parallel and Distributed Processing Symposium), Fort Lauderdale, FL, April.
13. Wu, J., Wyckoff, P., and Panda, D. K. 2003. Supporting efficient non-contiguous access in PVFS over InfiniBand. Proceedings of the IEEE International Conference on Cluster Computing, Hong Kong, China, December.
14. Wu, J., Wyckoff, P., and Panda, D. K. 2004. High performance implementation of MPI datatype communication over InfiniBand. International Parallel and Distributed Processing Symposium (IPDPS'04) , Santa Fe, NM, April.