[Table of Contents]

Plasma and Fusion Research

Volume 6, 2401069 (2011)

Regular Articles

High Performance Iterative Solver for Linear System using Multi GPU
Soichiro IKUNO1), Norihisa FUJITA1), Yuki KAWAGUCHI1), Taku ITOH2), Susumu NAKATA3), Kota WATANABE4) and Hiroyuki NAKAMURA5)
Tokyo University of Technology, 1404-1 Katakura, Hachioji, Tokyo 192-0982, Japan
Seikei University, 3-3-1 Kichijoji-kitamachi, Musashino, Tokyo 180-8633, Japan
Ritsumeikan University, 1-1-1 Nojihigashi, Kusatsu, Shiga 525-5062, Japan
Hokkaido University, 14-9 Kita-ku, Sapporo 060-0814, Japan
National Institute for Fusion Science, 322-6 Oroshi-cho, Toki, Gifu 509-5292, Japan
(Received 22 December 2010 / Accepted 28 March 2011 / Published 1 July 2011)


The variable preconditioned (VP) Krylov subspace method on multi Graphics Processing Unit (GPU) is numerically investigated. Besides, the linear system obtained by finite element method with an edge element is adopted for the problem. The results of computations show that VP conjugate gradient method on multi GPU demonstrated significant achievement than that of CPU. Especially, VP conjugate gradient method on multi GPU is 4.35 times faster than that of CPU. However, transmission rate between the PC using Gigabit Ethernet is the bottleneck of the performance.


variable preconditioned Krylov subspace method, GPGPU, CUDA, singular matrix, edge element

DOI: 10.1585/pfr.6.2401069


  • [1] S. Ikuno, M. Natori and A. Kamitani, Fusion Energy 1998, 4 (IAEA, Vienna, 1999) (Proc. 17th Int. Conf. on Plasma Physics and Contr. Nucl. Fusion Res., Yokohama, 1998) p.1497.
  • [2] http://gpgpu.org/
  • [3] A. Cevahir, A. Nukada and S. Matsuoka, Lecture Notes in Computer Science, 5544 (Springer Berlin, 2009) (Computational Science ICCS 2009) p.893.
  • [4] K. Abe and S.L. Zhang, Int. J. Numer. Anal. Model. 22, 147 (2005).
  • [5] http://www.nvidia.com/object/cuda_develop.html
  • [6] http://www.compumag.org/jsite/team.html
  • [7] H. Igarashi, IEEE Trans. Magn. 37, 3129 (2001).
  • [8] H. Igarashi and N. Yamamoto, IEEE Trans. Magn. 44, 942 (2008).
  • [9] K. Fujiwara, T. Nakata and H. Fusayasu, IEEE Trans. Magn. 29, 1958 (1993).

This paper may be cited as follows:

Soichiro IKUNO, Norihisa FUJITA, Yuki KAWAGUCHI, Taku ITOH, Susumu NAKATA, Kota WATANABE and Hiroyuki NAKAMURA, Plasma Fusion Res. 6, 2401069 (2011).