Abstract
This paper describes an efficient implementation of MPI on the Memory-Based Communication Facilities; Memory-Based FIFO is used for buffering by the library, and Remote Write for communication with no buffering. The Memory-Based Communication Facilities are software-based communication mechanisms, with off-the-shelf Ethernet hardware. They provide low-cost and highly-functional primitives for remote memory accesses.
The performance of the library was evaluated on a cluster of workstations connected with a 100Base-TX network. The round-trip time was 71 Μs for 0 byte message, and the peak bandwidth was 11.86 Mbyte/s in full-duplex mode. These values show that it is more efficient to realize the message passing libraries with the shared memory model than with the message passing model.
This work is partly supported by Information-technology Promotion Agency (IPA) Japan and by Real World Computing Partnership (RWCP) Japan.
This is a preview of subscription content, log in via an institution to check access.
Preview
Unable to display preview. Download preview PDF.
References
Message Passing Interface Forum. MPI: A Message-Passing Interface Standard. http://www.mpi-forum.org/, June 1995.
T. Matsumoto and K. Hiraki. Shared memory vs. message passing (in Japanese). In IPSJ SIG Reports 97-ARC-126, Vol. 97, No. 102, pages 85–90, October 1997.
T. Matsumoto, S. Furuso, and K. Hiraki. Resource management methods of the general-purpose massively-parallel operating system: SSS-CORE (in Japanese). In Proc. of 11th Conf. of JSSST, pages 13–16, October 1994.
T. Matsumoto, S. Uzuhara, and K. Hiraki. A general-purpose scalable operating system: SSS-CORE. In Proc. of 20th Int. Conf. on Software Engineering (2), pages 147–152, April 1998.
T. Matsumoto, T. Komaarashi, S. Uzuhara, S. Takeoka, and K. Hiraki. A general-purpose massively-parallel operating system: SSS-CORE — implementation methods for network of workstations-(in Japanese). In IPSJ SIG Notes 96-OS-73, Vol. 96, No. 79, pages 115–120, August 1996.
T. Matsumoto and K. Hiraki. MBCF: A protected and virtualized high-speed user-level memory-based communication facility. In Proc. of Int. Conf. Supercomputing '98 (to be appeared), July 1998.
H. Tezuka, A. Hori, Y. Ishikawa, and M. Sato. PM: An operating system coordinated high performance communication library. In B. Hertzberger and P. Sloot, editors, High-Performance Computing and Networking, volume 1225 of Lecture Notes in Computer Science, pages 708–717. Springer Verlag, April 1997.
W. Gropp, E. Lusk, N. Doss, and A. Skjellum. A high-performance, portable implementation of the MPI Message-Passing Interface Standard. Parallel Computing, 22(6):789–828, September 1996.
D. Sitsky and P. Mackerras. System developments on the Fujitsu AP3000. In P. Mackerras, editor, Proc. of 7th Parallel Computing Workshop, September 1997.
K. Cameron, L. Clarke, and G. Smith. CRI/EPCC MPI for CRAY T3D. http://www.epcc.ed.ac.uk/t3dmpi/Product/, September 1995.
O. Tatebe, Y. Kodama, S. Sekiguchi, and Y. Yamaguchi. Efficient implementation of MPI using remote memory write (in Japanese). In Proc. of Joint Symp. on Parallel Processing '98, pages 199–206, June 1998.
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 1998 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Morimoto, K., Matsumoto, T., Hiraki, K. (1998). Implementing MPI with the Memory-Based Communication Facilities on the SSS-CORE operating system. In: Alexandrov, V., Dongarra, J. (eds) Recent Advances in Parallel Virtual Machine and Message Passing Interface. EuroPVM/MPI 1998. Lecture Notes in Computer Science, vol 1497. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0056579
Download citation
DOI: https://doi.org/10.1007/BFb0056579
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-65041-6
Online ISBN: 978-3-540-49705-9
eBook Packages: Springer Book Archive