Efficiency Evaluation of PVM 2.X, PVM 3.X, P4, Express and Linda on a Workstation Cluster Using the NAS Parallel Benchmarks
When developing programs for a network of workstations, a tedious and time consuming task consists of handling the communication and synchronization in a correct and efficient way. Correct means in this context, that it is guaranteed, that each process running on a particular machine will receive the appropriate data and that the program will not deadlock. With the term efficient we mean a minimum amount of overhead needed for communication and synchronization, thus giving low execution times and high speedup.
A variety of tools have been developed supporting the user in this task. These tools include a parallelizing compiler, programming environments, and finally tools for message passing synchronization primitives. The developers of these tools claim, that their tools are easy to use, but will still produce efficient and correct code. But are all the tools equally efficient? To get an answer, a study has been performed evaluating five of the most commonly used tools, namely PVM 2.x, PVM 3.x, p4, Express (a registered Trademark of Parasoft Inc.) and Linda (a registered Trademark of Scientific Computing Associates Inc.) with respect to efficiency and also ease-of-use. Four kernels of the NAS parallel benchmarks have been selected for implementation on a workstation cluster. The results of this study are summarized in this paper.
Unable to display preview. Download preview PDF.
- D. Bailey, J. Barton, T. Lasinski, H. Simon, The NAS Parallel Benchmarks, Report RNR-91-002 Revision 2, Numerical Aerodynamic Simulation (NAS) Systems Division, NASA Ames Research Center, Moffett Field, August 1991.Google Scholar
- Jaswinder Pal Singh, Wolf-Dietrich Weber and Anoop Gupta, SPLASH: Stanford Parallel Applications for Shared-Memory, Computer Systems Laboratory, Stanford University, April 1991.Google Scholar
- A. Beguelin, J. Dongarra, Al Geist, R. Manchek, V. Sunderam, A USERS’ GUIDE TO PVM PARALLEL VIRTUAL MACHINE, ORNL/TM-11826, Oak Ridge National Laboratory, Dec 1991.Google Scholar
- A. Beguelin, J. Dongarra, Al Geist, W. Jiang, R. Manchek, V. Sunderam, PVM 3 USER’S GUIDE AND REFERENCE MANUAL, ORNL/TM-12187, Oak Ridge National Laboratory, May 1993.Google Scholar
- R. Butler, E. Lusk, User’s Guide to the p4 Programming System, ANL-92/17, ARGONNE NATIONAL LABORATORY, Oct 1992.Google Scholar
- Parasoft Corporation, Express C: User’s Guide 3.0, Reference 3.0, Release Notes 3.2, Pasadena, 1992.Google Scholar
- Scientific Computing Associates Inc., C-Linda User’s Guide & Reference Manual, New Haven, 1992.Google Scholar
- A. L. Beguelin, Xab: A Tool for Monitoring PVM Programs, Pittsburgh Supercomputer Center, Carnegie Mellon University, Pittsburgh.Google Scholar
- V. Herrarte, E. Lusk, Studying Parallel Behavior with Upshot.Google Scholar
- D. H. Bailey, E. Barszcz, J. T. Barton, D. S. Browing, R. L. Carter, L. Dagnum, R. A. Fatoohi, P. O. Frederickson, T. A. Lasinski, R. S. Schreiber, H. D. Simon, V. Venkatakrishnan and S. K. Weeratunga, The NAS Parallel Benchmarks — Summary and Preliminary Results, Proceedings of Super-computing 91, pp. 158–173, Nov 1991.Google Scholar
- High Performance Fortran Forum, High Performance Fortran Language Specification Version 1.0, Rice University, Houston Texas, May 1993.Google Scholar
- D. H. Bailey, E. Barszcz, L. Dagnum and H. D. Simon, NAS Parallel Benchmark Results 3–94, RNR Technical Report RNR-94-006, March 1994.Google Scholar
- J. J. Dongarra, Performance of Various Computers Using Standard Linear Equations Software, CS-89–85, University of Tennessee and Oak Ridge National Laboratory, Sep 1993.Google Scholar