A parallel computing method using blocked format with optimal partitioning for SpMV on GPU

作者：

Highlights：

• We develop an optimal partitioning strategy to improve the performance of SpMV.

• We present a reordering algorithm in which the time complexity is only .

• We employ a hybrid format to store a blocked sparse matrix partitioned by our optimal partitioning strategy.

摘要

•We develop an optimal partitioning strategy to improve the performance of SpMV.•We present a reordering algorithm in which the time complexity is only .•We employ a hybrid format to store a blocked sparse matrix partitioned by our optimal partitioning strategy.

论文关键词：Blocked format,CPU/GPU,Dynamic programming,Heterogeneous parallel computing,Partitioning,Reordering,Sparse matrix–vector multiplication

论文评审过程：Received 3 January 2016, Revised 9 June 2017, Accepted 26 September 2017, Available online 16 October 2017, Version of Record 13 November 2017.

论文官网地址：https://doi.org/10.1016/j.jcss.2017.09.010