Abstract
We present a perfectly balanced, "merge-based" parallel method for computing sparse matrix-vector products (SpMV). Our algorithm operates directly upon the Compressed Sparse Row (CSR) sparse matrix format, a predominant in-memory representation for general-purpose sparse linear algebra computations. Our CsrMV performs an equitable multi-partitioning of the input dataset, ensuring that no single thread can be overwhelmed by assignment to (a) arbitrarily-long rows or (b) an arbitrarily-large number of zero-length rows. This parallel decomposition requires neither offline preprocessing nor specialized/ancillary data formats. We evaluate our method on both CPU and GPU microarchitecture across an enormous corpus of diverse real world matrix datasets. We show that traditional CsrMV methods are inconsistent performers subject to order-of-magnitude slowdowns, whereas the performance response of our method is substantially impervious to row-length heterogeneity.
- T. Davis and Y. Hu. University of Florida Sparse Matrix Collection. URL http://www.cise.ufl.edu/research/sparse/matrices/. Google Scholar
Digital Library
- S. Filippone, V. Cardellini, D. Barbieri, and A. Fanfarillo. Parallel tridiagonal equation solvers. ACM Trans. Math. Softw. Google Scholar
Digital Library
- S. Odeh, O. Green, Z. Mwassi, O. Shmueli, and Y. Birk. Merge Path - Parallel Merging Made Simple. In Proceedings of the International Parallel and Distributed Processing Symposium Workshops and PhD Forum, pages 1611--1618, May 2012. Google Scholar
Digital Library
- R. W. Vuduc. Automatic Performance Tuning of Sparse Matrix Kernels. PhD thesis, 2003. AAI3121741. Google Scholar
Digital Library
- S. Williams, L. Oliker, R. Vuduc, J. Shalf, K. Yelick, and J. Demmel. Optimization of sparse matrix-vector multiplication on emerging multicore platforms. In Proceedings of the 2007 ACM/IEEE Conference on Supercomputing, SC '07, pages 38:1--38:12, New York, NY, USA, 2007. ACM. ISBN 978-1-59593-764-3. doi: 10.1145/1362622. 1362674. Google Scholar
Digital Library
Index Terms
Merge-based sparse matrix-vector multiplication (SpMV) using the CSR storage format
Recommendations
Merge-based sparse matrix-vector multiplication (SpMV) using the CSR storage format
PPoPP '16: Proceedings of the 21st ACM SIGPLAN Symposium on Principles and Practice of Parallel ProgrammingWe present a perfectly balanced, "merge-based" parallel method for computing sparse matrix-vector products (SpMV). Our algorithm operates directly upon the Compressed Sparse Row (CSR) sparse matrix format, a predominant in-memory representation for ...
Merge-based parallel sparse matrix-vector multiplication
SC '16: Proceedings of the International Conference for High Performance Computing, Networking, Storage and AnalysisWe present a strictly balanced method for the parallel computation of sparse matrix-vector products (SpMV). Our algorithm operates directly upon the Compressed Sparse Row (CSR) sparse matrix format without preprocessing, inspection, reformatting, or ...
Optimizing Sparse Matrix Vector Multiplication Using Diagonal Storage Matrix Format
HPCC '10: Proceedings of the 2010 IEEE 12th International Conference on High Performance Computing and CommunicationsSparse matrix vector multiplication (SpMV) is used in many scientific computations. The main bottleneck of this algorithm is memory bandwidth and many methods reduce memory bandwidth usage by compressing the index array. The matrices from finite ...






Comments