Abstract
A need for a more efficient programming paradigm has been prompted by the introduction of networks of symmetric multiprocessor (SMP) workstations. A new programming paradigm for networks of SMP workstations is presented in this paper. The paradigm called MPIT integrates Message Passing Interface (MPI) and POSIX threads. The MPIT paradigm utilizes MPI for communication among the workstations, and uses threads to process the data within a workstation. The communication among the workstations is handled by a dedicated communication thread that runs on each workstation. The communication among the threads is handled through the shared memory. A number of theoretical and practical benefits of the MPIT paradigm are identified, such as communication/computation overlap, increased resource utilization and performance.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
TOP500 Supercomputer Sites, University of Mannheim and University of Tennessee, http://www.top500.org.
MPI Forum: MPI-1.1 Standard. http://www.mpi-forum.org (1995).
MPI Forum: MPI-2 Standard. http://www.mpi-forum.org (1997).
Sunderam V.S.: PVM: A Framework for Parallel Distributed Computing. Concurrency: Theory and Practice 2 (1990) 315–339.
Tang, H., Shen, K., Yang, T.: Compile/Run-time Support for Threaded MPI Execution on Multiprogrammed Shared Memory Machines. Proceedings of ACM Programming Principles of Parallel Processing (1999) 107–118.
Protopopov, B.V., Skjellum, A.: A multi-threaded Message Passing Interface (MPI) architecture: performance and program issues. Journal of Parallel and Distributed Computing 61 (2001) 449–466.
Haines, M., Cronk, D., Mehrotra, P.: On the Design of Chant: A Talking Threads Package. Proceedings of the Conference on Supercomputing (1994) 350–359.
Tanaka, Y., Matsuda, M., Kubota, K., Sato, M.: COMPaS: A Pentium Pro PC-Based SMP Cluster. High Performance Cluster Computing, Volume 1. 1st edition. Prentice Hall (1999) 661–681.
Parab, N., Raghvedran, M.: Active Messages. High Performance Cluster Computing, Volume 1. 1st edition. Prentice Hall (1999) 270–300.
Bader, D.A., Jájá, J.: SIMPLE: A Methodology for Programming High Performance Algorithms on Clusters of Symmetric Multiprocessors (SMPs). Journal of Parallel and Distributed Computing 58 (1999) 92–108.
Butenhof, D.R.: Programming with POSIX Threads. Addison-Wesley, 1997, ISBN 0-201-64492-2.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2002 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Huttunen, P., Ikonen, J., Porras, J. (2002). MPIT — Communication/Computation Paradigm for Networks of SMP Workstations. In: Fagerholm, J., Haataja, J., Järvinen, J., Lyly, M., Råback, P., Savolainen, V. (eds) Applied Parallel Computing. PARA 2002. Lecture Notes in Computer Science, vol 2367. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-48051-X_17
Download citation
DOI: https://doi.org/10.1007/3-540-48051-X_17
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-43786-4
Online ISBN: 978-3-540-48051-8
eBook Packages: Springer Book Archive