Skip to main content
SpringerLink
Log in
Book cover

International Workshop on Applied Parallel Computing

PARA 1994: Parallel Scientific Computing pp 67–79Cite as

A transform approach to fast matrix multiplication

  • Conference paper
  • First Online:

Part of the book series: Lecture Notes in Computer Science ((LNCS,volume 879))

Abstract

The Strassen fast matrix multiplication algorithm and other related recursive fast matrix multiplication algorithms can be re-cast in terms of a matrix transform approach which is conceptually similar to the Fast Fourier Transform. This approach allows many alternate but equivalent formulations of the Strassen-like algorithms through addressing algebra. We give examples of how such techniques can be applied to a fine-grained massively parallel system (the MasPar MP-2), a cache-based system and the out-of-core matrix multiplication problem and the implications for extra memory requirements as well as data movement.

This is a preview of subscription content, log in via an institution.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. V. Strassen (1969), Gaussian elimination is not optimal, Numer. Math, v.13, pp. 354–356.

    Google Scholar 

  2. S. Winograd (1968), A new algorithm for inner product, IEEE Trans, on Computing, v. C-18. pp. 693–694.

    Google Scholar 

  3. V. Pan (1984), How can we speed up matrix multiplication?, SIAM Review, v.26, No.3, pp. 393–415.

    Google Scholar 

  4. V. Pan (1980), New fast algorithms for matrix operations, SIAM J. Comput., v.9, No. 2, pp. 321–342.

    Google Scholar 

  5. D. Coppersmith, S. Winograd (1990), Matrix multiplication via arithmetic progression, J. Symbolic Comp., v.9, pp. 251–280.

    Google Scholar 

  6. David Bailey (1988), Extra high speed matrix multiplication on the CRAY-2, SIAM J. Sci.Stat. Comput., 9, No.3, pp. 603–607.

    Google Scholar 

  7. D.H. Bailey, K. Lee, H.D. Simon (1990), Using Strassen's algorithn to accelerate the solution of linear systems. J. Supercomp., v. 4, pp.357–371.

    Google Scholar 

  8. P. Bjorstad, F. Manne, T. Sorevik, M Vajtersic (1991), Efficient matrix multiplication on SIMD computers, Institutt for Informatikk, Univ. of Bergen, Report No. 53.

    Google Scholar 

  9. G. Golub, J. Ortega (1993), Scientific Computing. An Introduction with Parallel Computing. Academic Press.

    Google Scholar 

  10. S.M. Balle, P.C. Hansen, N.J. Higham (1993), A Strassen-Type Matrix Inversion Algorithm for the Connection Machine, Report UNIC-93-11.

    Google Scholar 

  11. W. Miller (1975), Computational complexity and numerical stability, SIAM J. Comput, v. 4, No.3, pp. 97–107.

    Google Scholar 

  12. N.J. Higham (1990), Exploiting fast matrix multiplication within the level 3 BLAS, ACM Trans. Math. Software, v. 16, pp.352–368.

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. 1045 Forest Knoll Dr., 95129, San Jose, CA, USA

    Pavel Berkhin

  2. Pentum Group, Inc., 485 Central Ave., 94043, Mountain View, CA, USA

    John Brown

Authors
  1. Pavel Berkhin
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. John Brown
    View author publications

    You can also search for this author in PubMed Google Scholar

Editor information

Jack Dongarra Jerzy Waśniewski

Rights and permissions

Reprints and permissions

Copyright information

© 1994 Springer-Verlag Berlin Heidelberg

About this paper

Cite this paper

Berkhin, P., Brown, J. (1994). A transform approach to fast matrix multiplication. In: Dongarra, J., Waśniewski, J. (eds) Parallel Scientific Computing. PARA 1994. Lecture Notes in Computer Science, vol 879. Springer, Berlin, Heidelberg. https://doi.org/10.1007/BFb0030137

Download citation

  • DOI: https://doi.org/10.1007/BFb0030137

  • Published:

  • Publisher Name: Springer, Berlin, Heidelberg

  • Print ISBN: 978-3-540-58712-5

  • Online ISBN: 978-3-540-49050-0

  • eBook Packages: Springer Book Archive

Publish with us

Policies and ethics

Search

Navigation