Advertisement

Reduction to Condensed Forms for Symmetric Eigenvalue Problems on Multi-core Architectures

  • Paolo Bientinesi
  • Francisco D. Igual
  • Daniel Kressner
  • Enrique S. Quintana-Ortí
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6067)

Abstract

We investigate the performance of the routines in LAPACK and the Successive Band Reduction (SBR) toolbox for the reduction of a dense matrix to tridiagonal form, a crucial preprocessing stage in the solution of the symmetric eigenvalue problem, on general-purpose multi-core processors. In response to the advances of hardware accelerators, we also modify the code in SBR to accelerate the computation by off-loading a significant part of the operations to a graphics processor (GPU). Performance results illustrate the parallelism and scalability of these algorithms on current high-performance multi-core architectures.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Golub, G.H., Loan, C.F.V.: Matrix Computations, 3rd edn. The Johns Hopkins University Press, Baltimore (1996)zbMATHGoogle Scholar
  2. 2.
    Dhillon, S., Parlett, N.: Vomel: The design and implementation of the MRRR algorithm. ACM Trans. Math. Soft. 32(4), 533–560 (2006)CrossRefMathSciNetGoogle Scholar
  3. 3.
    Bientinesi, P., Dhillon, I.S., van de Geijn, R.: A parallel eigensolver for dense symmetric matrices based on multiple relatively robust representations. SIAM J. Sci. Comput. 27(1), 43–66 (2005)zbMATHCrossRefMathSciNetGoogle Scholar
  4. 4.
    Bischof, C.H., Lang, B., Sun, X.: Algorithm 807: The SBR Toolbox—software for successive band reduction. ACM Trans. Math. Soft. 26(4), 602–616 (2000)CrossRefMathSciNetGoogle Scholar
  5. 5.
    Dongarra, J., Hammarling, S.J., Sorensen, D.C.: Block reduction of matrices to condensed forms for eigenvalue computations. LAPACK Working Note 2, Technical Report MCS-TM-99, Argonne National Laboratory (September 1987)Google Scholar
  6. 6.
    Bientinesi, P., Igual, F.D., Kressner, D., Quintana-Ortí, E.S.: Reduction to condensed forms for symmetric eigenvalue problems on multi-core architectures. Technical Report 2009-13, Seminar for Applied Mathematics, ETH Zurich, Switzerland (2009), http://www.sam.math.ethz.ch/reports/2009
  7. 7.
    Barrachina, S., Castillo, M., Igual, F.D., Mayo, R., Quintana-Ortí, E.S.: Evaluation and tuning of the level 3 CUBLAS for graphics processors. In: Proceedings of the 10th IEEE Workshop on Parallel and Distributed Scientific and Engineering Computing, PDSEC 2008 (2008), CD-ROMGoogle Scholar
  8. 8.
    Barrachina, S., Castillo, M., Igual, F.D., Mayo, R., Quintana-Ortí, E.S.: Solving dense linear systems on graphics processors. In: Luque, E., Margalef, T., Benítez, D. (eds.) Euro-Par 2008. LNCS, vol. 5168, pp. 739–748. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  9. 9.
    Volkov, V., Demmel, J.W.: Benchmarking gpus to tune dense linear algebra. In: SC 2008: Proceedings of the 2008 ACM/IEEE Conference on Supercomputing, Piscataway, NJ, USA, pp. 1–11. IEEE Press, Los Alamitos (2008)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Paolo Bientinesi
    • 1
  • Francisco D. Igual
    • 2
  • Daniel Kressner
    • 3
  • Enrique S. Quintana-Ortí
    • 2
  1. 1.AICES, RWTH Aachen UniversityAachenGermany
  2. 2.Depto. de Ingeniería y Ciencia de ComputadoresUniversidad Jaume ICastellónSpain
  3. 3.Seminar für angewandte MathematikETH ZürichSwitzerland

Personalised recommendations