Advertisement

Embedding SCI into PVM

  • Markus Fischer
  • Jens Simon
3 Implementation Issues
Part of the Lecture Notes in Computer Science book series (LNCS, volume 1332)

Abstract

The extremely low latencies and high bandwidth results achievable with the Scalable Coherent Interface (SCI) at lowest level encourages its integration into existing Message Passing Environments (MPE's). In combination with Network of Workstations, it can be seen as an alternative for traditional parallel computing with tightly coupled processors with comparable performance at low cost. This paper describes the ongoing implementation of PVM using SCI hardware devices and with Linux and WindowsNT as the operating systems. It gives an overview of SCI, its performance and the possibilities for PVM to make use of the superior features of SCI compared with the conventional Ethernet.

Keywords

Virtual Machine Shared Memory Message Passing Interface Message Passing Remote Memory 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    Hellwagner et al. A PCI-SCI bridge for building a PC-cluster with distributed shared memory. In 6th Int. Workshop on SCI-based High-Performance Low-Cost Computing — SCIzzL, September 1996.Google Scholar
  2. [2]
    Convex SPP S-Class and X-Class Systems Overview. Convex Division of Hewlett-Packard, 1996.Google Scholar
  3. [3]
    J. Dongarra and M. Fischer. PVM on Windows 95/NT. In Annual CRPC meeting, Chicago, May 1996.Google Scholar
  4. [4]
    PCI-SCI Adapter Card Specification and Modular Switch. Dolphin Interconnect Solutions, Inc., Norway, '96/'97.Google Scholar
  5. [5]
    Dormanns et al. A programming interface for NUMA shared-memory clusters. High Perf. Computing and Networking (HPCN'97), April 1997.Google Scholar
  6. [6]
    J. Dongarra et al. PVM: Parallel Virtual Machine. A User's Guide and Tutorial for Networked Parallel Computing. MIT Press, Boston, 1994.Google Scholar
  7. [7]
    A. Geist, J. Kohl and P. Papadopoulos. PVM and MPI: A Comparison of Features. Published in Calculateurs Paralleles Vol. 8 No. 2 (1996).Google Scholar
  8. [8]
    IEEE Std for Scalable Coherent Interface (SCI). Inst. of Electrical and Electronical Eng., Inc., New York, NY 10017, IEEE std 1596–1992, 1993.Google Scholar
  9. [9]
    Ramme et al. A Distrib. Comp. Center Software for the Efficient Use of Par. Computer Systems. In HPCN'94, LNCS 797, pages 129–136, 1994.Google Scholar
  10. [10]
    Sequent's NUMA-Q 2000 Solutions: Reliability, Availability, Scalability. Sequent Computer Systems, 1997.Google Scholar
  11. [11]
    J. Simon and O. Heinz. Experiences with a SCI multiprocessor workstation cluster. In Architecture of Computer Systems, ARCS'97, 1997.Google Scholar
  12. [12]
    J. Weidendoerfer Entwurf und Implementierung einer Socket Bibliothek fuer ein SCI Netzwerk Master Thesis, University of Munich (TUM), 1997.Google Scholar
  13. [13]
    H. Zhou and A. Geist Faster Message Passing in PVM. Technical Report, Oak Ridge National Laboratory, Oak Ridge, TN 37831, 1995.Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 1997

Authors and Affiliations

  • Markus Fischer
    • 1
  • Jens Simon
    • 1
  1. 1.PC2 - Paderborn Center for Parallel ComputingUSA

Personalised recommendations