sparsematrix operations are frequently used operations in scientific, engineering and high-performance computing (HPC) applications. Among them, sparsematrix-vectormultiplication (SpMV) is a popular kernel and cons...
详细信息
sparsematrix operations are frequently used operations in scientific, engineering and high-performance computing (HPC) applications. Among them, sparsematrix-vectormultiplication (SpMV) is a popular kernel and considered an important numerical method for science, engineering and in scientific computing. However, SpMV is a computationally expensive operation. To obtain better performance, SpMV depends on certain factors;choosing the right storage format for the sparsematrix is one of them. Other things like data access pattern, the sparsity of the matrix data set, load balancing, sharing of the memory hierarchy, etc. are other factors that affect performance. Metadata, that describes the substructure of the sparsematrix, like shape, density, sparsity, etc. of the sparsematrix also affects performance efficiency for any sparsematrix operation. Various approaches presented in literature over the last few decades given good results for certain types of matrix structures and don't perform as well with others. Developers thus are faced with a difficulty in choosing the most appropriate format. In this research, an approach is presented that evaluates metadata of a given sparsematrix and suggest to the developers the most suitable storage format to use for SpMV.
暂无评论