Advertisement

A Parallel Variant of BiCGStar-Plus Method Reduced to Single Global Synchronization

  • Seiji Fujino
  • Keiichi Murakami
Part of the Communications in Computer and Information Science book series (CCIS, volume 402)

Abstract

In this paper, we propose new product-type iterative methods by introducing the BiCGSafe strategy, i.e., utilization of associate residual in place of residual, to the variants of GPBiCG. We refer to BiCGStar (BiCG with stabilization of associate residual) and its improved version of BiCGStar method. BiCGStar and BiCGStar-plus methods outperform compared with the conventional iterative methods. Moreover, our proposed methods are suited to parallel computer with distributed memory systems, since they require single global synchronization per one iteration.

Keywords

iterative method synchronization parallel computer 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Abe, K., Sleijpen, G.L.G.: Solving linear equations with a stabilized GPBiCG mehtod. Appl. Numer. Math. 67, 4–16 (2013)MathSciNetCrossRefzbMATHGoogle Scholar
  2. 2.
    Abe, K., Sleijpen, G.L.G.: Hybrid Bi-CG methods with Bi-CG formulation closer to the IDR approach. Appl. Math. Comput. 218, 10889–10899 (2012)MathSciNetCrossRefGoogle Scholar
  3. 3.
  4. 4.
    Fujino, S., Fujiwara, M., Yoshida, M.: A proposal of preconditioned BiCGSafe method with safe convergence. In: Proc. of the 17th IMACS World Congress on Scientific Computation, Appl. Math. Simul. CD-ROM, Paris (2005)Google Scholar
  5. 5.
    Rutishauser, H.: Theory of gradient method, Refined Iterative Methods for Computation of the Solution and the Eigenvalues of Self-Adjoint Boundary Value Problems. Mitt. Inst. Angew. Math, pp. 24–49. ETH Zürich, Birkhäuser (1959)Google Scholar
  6. 6.
    Sleijpen, G.L.G., Sonneveld, P., van Gijzen, M.B.: Bi-CGSTAB as an induced dimension reduction method. Appl. Numer. Math. 60, 1100–1114 (2010)Google Scholar
  7. 7.
    Sonneveld, P., van Gijzen, M.B.: IDR(s): a family of simple and fast algorithms for solving large nonsymmetric linear systems. SIAM J. Sci. Stat. Comput. 31(2), 1035–1062 (2008)CrossRefzbMATHGoogle Scholar
  8. 8.
    Zhang, S.-L.: GPBi-CG: Generalized product-type methods based on Bi-CG for solving nonsymmetric linear systems. SIAM J. Sci. Comput. 18, 537–551 (1997)MathSciNetCrossRefzbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2013

Authors and Affiliations

  • Seiji Fujino
    • 1
  • Keiichi Murakami
    • 2
  1. 1.Research Institute for Information TechnologyKyushu UniversityJapan
  2. 2.Sumitomo Rubber Industries, Ltd.Japan

Personalised recommendations