Refine your search:     
Report No.

Implementation and performance evaluation of a communication-avoiding GMRES method for stencil-based code on GPU cluster

Matsumoto, Kazuya*; Idomura, Yasuhiro; Ina, Takuya*; Mayumi, Akie; Yamada, Susumu

A communication-avoiding generalized minimum residual method (CA-GMRES) is implemented on a hybrid CPU-GPU cluster, targeted for the performance acceleration of iterative linear system solver in the gyrokinetic toroidal five-dimensional Eulerian code GT5D. In addition to the CA-GMRES, we implement and evaluate a modified variant of CA-GMRES (M-CA-GMRES) proposed in our previous study to reduce the amount of floating-point calculations. This study demonstrates that beneficial features of the CA-GMRES are in its minimum number of collective communications and its highly efficient calculations based on dense matrix-matrix operations. The performance evaluation is conducted on the Reedbush-L GPU cluster, which contains four NVIDIA Tesla P100 GPUs per compute node. The evaluation results show that the M-CA-GMRES is 1.09x, 1.22x and 1.50x faster than the CA-GMRES, the generalized conjugate residual method (GCR), and the GMRES, respectively, when 64 GPUs are used.



- Accesses




Category:Computer Science, Hardware & Architecture



[CLARIVATE ANALYTICS], [WEB OF SCIENCE], [HIGHLY CITED PAPER & CUP LOGO] and [HOT PAPER & FIRE LOGO] are trademarks of Clarivate Analytics, and/or its affiliated company or companies, and used herein by permission and/or license.