Plasma and Fusion Research
Volume 5, S2115 (2010)
Regular Articles
- Tokyo University of Technology, Katakura 1404-1, Hachioji, Tokyo 192-0982, Japan
Abstract
The Variable Preconditioned GCR (VPGCR) with mixed precision on Graphics Processing Unit (GPU) using CUDA is numerically investigated. The convergence theorem of VPGCR is guaranteed that the residual equation can be solved in the range of single precision, which means that VPGCR is applicable method to elicit the high performance of GPU. The results of computations show that VPGCR with mixed precision demonstrated significant achievement than that with double precision operation. Especially, the hybrid VPGCR on GPU is 10.10 times faster than that of CPU, and the standard VPGCR on GPU is 10.36 times faster than that of CPU.
Keywords
VPGCR method, GPGPU, CUDA, mixed precision, linear system
Full Text
References
- [1] S. Ikuno, M. Natori and A. Kamitani, Fusion Energy 1998, 4, 1497 (IAEA, Vienna, 1999).
- [2] http://gpgpu.org/
- [3] A. Cevahir, A. Nukada and S. Matsuoka, Lecture Notes in Computer Science, 5544, 893 (Springer Berlin, 2009).
- [4] K. Abe and S. L. Zhang, Int. J. Numer. Anal. Model. 22, 147 (2005).
- [5] http://www.nvidia.com/object/cuda develop.html
This paper may be cited as follows:
Soichiro IKUNO, Shuntaro DEKI and Norihisa FUJITA, Plasma Fusion Res. 5, S2115 (2010).