ISCAS OpenIR  > 并行软件与计算科学实验室 
crsd: application specific auto-tuning of spmv for diagonal sparse matrices
Sun Xiangzheng; Zhang Yunquan; Wang Ting; Long Guoping; Zhang Xianyi; Li Yan
2011
Conference Name17th International Conference on Parallel Processing, Euro-Par 2011
SourceLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages316-327
Conference DateAugust 29,
Conference PlaceBordeaux, France
Publish PlaceGermany
ISSN3029743
ISBN9783642233968
Department(1) Lab. of Parallel Software and Computational Science, Institute of Software, Graduate University of Chinese Academy of Sciences, China
English AbstractSparse Matrix-Vector multiplication (SpMV) is an important computational kernel in scientific applications. Its performance highly depends on the nonzero distribution of sparse matrices. In this paper, we propose a new storage format for diagonal sparse matrices, defined as Compressed Row Segment with Diagonal-pattern (CRSD). We design diagonal patterns to represent the diagonal distribution. As the diagonal distributions are similar within matrices from one application, some diagonal patterns remain unchanged. First, we sample one matrix to obtain the unchanged diagonal patterns. Next, the optimal SpMV codelets are generated automatically for those diagonal patterns. Finally, we combine the generated codelets as the optimal SpMV implementation. In addition, the information collected during auto-tuning process is also utilized for parallel implementation to achieve load-balance. Experimental results demonstrate that the speedup reaches up to 2.37 (1.70 on average) in comparison with DIA and 4.60 (2.10 on average) in comparison with CSR under the same number of threads on two mainstream multi-core platforms. © 2011 Springer-Verlag.
KeywordDistributed Computer Systems Optimization
Content Type会议论文
URIhttp://ir.iscas.ac.cn/handle/311060/14335
Collection并行软件与计算科学实验室 
Recommended Citation
GB/T 7714
Sun Xiangzheng,Zhang Yunquan,Wang Ting,et al. crsd: application specific auto-tuning of spmv for diagonal sparse matrices[C]. Germany,2011:316-327.
Files in This Item:
File Name/Size DocType Version Access License
crsd application spe(606KB) 开放获取--Application Full Text
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Sun Xiangzheng]'s Articles
[Zhang Yunquan]'s Articles
[Wang Ting]'s Articles
Baidu academic
Similar articles in Baidu academic
[Sun Xiangzheng]'s Articles
[Zhang Yunquan]'s Articles
[Wang Ting]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Sun Xiangzheng]'s Articles
[Zhang Yunquan]'s Articles
[Wang Ting]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.