Title :
Auto-Tuning CUDA Parameters for Sparse Matrix-Vector Multiplication on GPUs
Author :
Guo, Ping ; Wang, Liqiang
Author_Institution :
Dept. of Comput. Sci., Univ. of Wyoming, Laramie, WY, USA
Abstract :
Graphics Processing Unit (GPU) has become an attractive coprocessor for scientific computing due to its massive processing capability. The sparse matrix-vector multiplication (SpMV) is a critical operation in a wide variety of scientific and engineering applications, such as sparse linear algebra and image processing. This paper presents an auto-tuning framework that can automatically compute and select CUDA parameters for SpMV to obtain the optimal performance on specific GPUs. The framework is evaluated on two NVIDIA GPU platforms, GeForce 9500 GTX and GeForce GTX 295.
Keywords :
coprocessors; matrix multiplication; sparse matrices; tuning; GeForce 9500 GTX; GeForce GTX 295; NVIDIA GPU platforms; SpMV; auto-tuning CUDA parameters; auto-tuning framework; coprocessor; graphics processing unit; sparse matrix-vector multiplication; Finite element methods; Graphics processing unit; Instruction sets; Kernel; Performance evaluation; Sparse matrices; Tuning; CUDA; GPU; performance; sparse matrix-vector multiplication;
Conference_Titel :
Computational and Information Sciences (ICCIS), 2010 International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-8814-8
Electronic_ISBN :
978-0-7695-4270-6
DOI :
10.1109/ICCIS.2010.285