Skip to main content

Model-driven autotuning of sparse matrix-vector multiply on GPUs

JeeWhan Choi, Amik Singh, Richard W. Vuduc

DOI 10.1145/1693453.1693471 dblp BibTeX

Abstract

We present a performance model-driven framework for automated performance tuning (autotuning) of sparse matrix-vector multiply (SpMV) on systems accelerated by graphics processing units (GPU). Our study consists of two parts.

Related papers

VDHA: Vector-Driven Hash Aggregation for Sparse Matrix-Sparse Vector Multiplication on GPUs PPoPP 2026
Adaptive sparse matrix-matrix multiplication on the GPU PPoPP 2019
Optimization by runtime specialization for sparse matrix-vector multiplication GPCE 2014
Merge-based sparse matrix-vector multiplication (SpMV) using the CSR storage format PPoPP 2016
Generating fast sparse matrix vector multiplication from a high level generic functional IR CC 2020
Dynamic N: M Fine-Grained Structured Sparse Attention Mechanism PPoPP 2023
Generating piecewise-regular code from irregular structures PLDI 2019
yaSpMV: yet another SpMV framework on GPUs PPoPP 2014