Paper Info
Title
Improving the Performance of Multithreaded Sparse Matrix-Vector Multiplication Using Index and Value Compression
Abstract
The Sparse Matrix-Vector Multiplication kernel exhibits limited potential for taking advantage of modern shared memory architectures due to its large memory bandwidth requirements. To decrease memory contention and improve the performance of the kernel we propose two compression schemes. The first, called CSR-DU, targets the reduction of the matrix structural data by applying coarse grain delta encoding for the column indices. The second scheme, called CSR-VI, targets the reduction of the numerical values using indirect indexing and can only be applied to matrices which contain a small number of unique values. Evaluation of both methods on a rich matrix set showed that they can significantly improve the performance of the multithreaded version of the kernel and achieve good scalability for large matrices.
Year
DOI
Venue
2008
10.1109/ICPP.2008.62
ICPP
Keywords
Field
DocType
value compression,column index,large matrix,multithreaded sparse matrix-vector multiplication,sparse matrix-vector multiplication kernel,memory contention,rich matrix set,modern shared memory,coarse grain delta,matrix structural data,compression scheme,large memory bandwidth requirement,kernel,memory bandwidth,multi threading,vectors,matrix multiplication,structured data,artificial neural networks,sparse matrices,sparse matrix,shared memory,indexes
Kernel (linear algebra),Memory bandwidth,Shared memory,Sparse matrix-vector multiplication,Computer science,Matrix (mathematics),Parallel computing,Matrix multiplication,Memory architecture,Sparse matrix
Conference
Citations 
PageRank 
References 
17
1.03
20
Authors
3
Name
Order
Citations
PageRank
Kornilios Kourtis134029.44
Georgios Goumas226822.03
N. Koziris31015107.53