Load-balancing Sparse Matrix Vector Product Kernels on GPUs

Citations Over TimeTop 10% of 2020 papers

Abstract

Efficient processing of Irregular Matrices on Single Instruction, Multiple Data (SIMD)-type architectures is a persistent challenge. Resolving it requires innovations in the development of data formats, computational techniques, and implementations that strike a balance between thread divergence, which is inherent for Irregular Matrices, and padding, which alleviates the performance-detrimental thread divergence but introduces artificial overheads. To this end, in this article, we address the challenge of designing high performance sparse matrix-vector product (S p MV) kernels designed for Nvidia Graphics Processing Units (GPUs). We present a compressed sparse row (CSR) format suitable for unbalanced matrices. We also provide a load-balancing kernel for the coordinate (COO) matrix format and extend it to a hybrid algorithm that stores part of the matrix in SIMD-friendly Ellpack format (ELL) format. The ratio between the ELL- and the COO-part is determined using a theoretical analysis of the nonzeros-per-row distribution. For the over 2,800 test matrices available in the Suite Sparse matrix collection, we compare the performance against S p MV kernels provided by NVIDIA’s cuSPARSE library and a heavily-tuned sliced ELL (SELL-P) kernel that prevents unnecessary padding by considering the irregular matrices as a combination of matrix blocks stored in ELL format.

Related Papers

• → Sparse Matrix Multiplication On An Associative Processor(2014)35 cited
• → Implementing Sparse Matrix-Vector multiplication using CUDA based on a hybrid sparse matrix format(2010)24 cited
• → Implementation of selected numerical algorithms for solving sparse matrixes using CUDA technology(2013)6 cited
• → SpMacho - Optimizing Sparse Linear Algebra Expressions with Probabilistic Density Estimation(2015)7 cited
• Data Regroup and Alignment Optimization Based on SIMD(2012)