SMAT: An Input Adaptive Sparse Matrix-Vector Multiplication Auto-Tuner

TL;DR

SMAT is an auto-tuning framework that automatically selects the optimal sparse matrix format and implementation at runtime, significantly improving performance for sparse matrix-vector multiplication across different hardware and matrix types.

Contribution

It introduces a unified auto-tuner that dynamically chooses the best sparse matrix format and implementation, bridging the gap between optimized kernels and general-purpose use.

Findings

SMAT achieves up to 75 GFLOP/s in single-precision on Intel.

SMAT outperforms MKL sparse functions by over 3 times.

The data mining model effectively selects optimal formats based on 2373 matrices.

Abstract

Sparse matrix vector multiplication (SpMV) is an important kernel in scientific and engineering applications. The previous optimizations are sparse matrix format specific and expose the choice of the best format to application programmers. In this work we develop an auto-tuning framework to bridge gap between the specific optimized kernels and their general-purpose use. We propose an SpMV auto-tuner (SMAT) that provides an unified interface based on compressed sparse row (CSR) to programmers by implicitly choosing the best format and the fastest implementation of any input sparse matrix in runtime. SMAT leverage a data mining model, which is formulated based on a set of performance parameters extracted from 2373 matrices in UF sparse matrix collection, to fast search the best combination. The experiments show that SMAT achieves the maximum performance of 75 GFLOP/s in single-precision and 33 GFLOP/s in double-precision on Intel, and 41 GFLOP/s in single-precision and 34 GFLOP/s in double-precision on AMD. Compared with the sparse functions in MKL library, SMAT runs faster by more than 3 times.