Analysis of Floating-Point Matrix Multiplication Computed via Integer Arithmetic

Abstract. Ootomo, Ozaki, and Yokota [ Int. J. High Perform. Comput. Appl., 38 (2024), pp. 297–313] have proposed a strategy to recast a floating-point matrix multiplication in terms of integer matrix products. The factors [Formula: see text] and [Formula: see text] are split into integer slices, the product of these slices is computed exactly, and [Formula: see text] is approximated by accumulating these integer products in floating-point arithmetic. This technique is particularly well suited to mixed-precision matrix multiply–accumulate units with integer support, such as the NVIDIA tensor cores or the AMD matrix cores. The number of slices allows for performance-accuracy tradeoffs: more slices yield better accuracy but require more multiplications, which in turn reduce performance. We propose an inexpensive way to estimate the minimum number of multiplications needed to achieve a prescribed level of accuracy. Our error analysis shows that the algorithm may become inaccurate (or inefficient) if rows of [Formula: see text] or columns of [Formula: see text] are badly scaled. We perform a range of numerical experiments, both in simulation and on the latest NVIDIA GPUs, that confirm the analysis and illustrate strengths and weaknesses of the algorithm. Reproducibility of computational results. This paper has been awarded the “SIAM Reproducibility Badge: Code and data available” as a recognition that the authors have followed reproducibility principles valued by SISC and the scientific computing community. Code and data that allow readers to reproduce the results in this paper are available at https://github.com/north-numerical-computing/integer-matrix-multiply-experiments and in the supplementary materials ( integer-matrix-multiply-experiments-main.zip [97.8KB]). [Formula: see text]

Authors

Institutions

Publication Details

Journal
SIAM Journal on Scientific Computing
Published
2026-11-16
DOI
https://doi.org/10.1137/25m1768539
Citations
19
Primary Topic
Numerical Methods and Algorithms
Type
article
Field-Weighted Citation Impact
27.90

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Analysis of Floating-Point Matrix Multiplication Computed via Integer Arithmetic

Massimiliano Fasi, Mantas Mikaitis, Françoise Tisseur B, Jack J. Dongarra et al.
19 citations
SIAM Journal on Scientific Computing
Numerical Methods and Algorithms
27.90
article

Analysis of Floating-Point Matrix Multiplication Computed via Integer Arithmetic

Massimiliano Fasi, Mantas Mikaitis, Françoise Tisseur B, Jack J. Dongarra, Ahmad Abdelfattah
article en
19 citations

Abstract

Abstract. Ootomo, Ozaki, and Yokota [ Int. J. High Perform. Comput. Appl., 38 (2024), pp. 297–313] have proposed a strategy to recast a floating-point matrix multiplication in terms of integer matrix products. The factors [Formula: see text] and [Formula: see text] are split into integer slices, the product of these slices is computed exactly, and [Formula: see text] is approximated by accumulating these integer products in floating-point arithmetic. This technique is particularly well suited to mixed-precision matrix multiply–accumulate units with integer support, such as the NVIDIA tensor cores or the AMD matrix cores. The number of slices allows for performance-accuracy tradeoffs: more slices yield better accuracy but require more multiplications, which in turn reduce performance. We propose an inexpensive way to estimate the minimum number of multiplications needed to achieve a prescribed level of accuracy. Our error analysis shows that the algorithm may become inaccurate (or inefficient) if rows of [Formula: see text] or columns of [Formula: see text] are badly scaled. We perform a range of numerical experiments, both in simulation and on the latest NVIDIA GPUs, that confirm the analysis and illustrate strengths and weaknesses of the algorithm. Reproducibility of computational results. This paper has been awarded the “SIAM Reproducibility Badge: Code and data available” as a recognition that the authors have followed reproducibility principles valued by SISC and the scientific computing community. Code and data that allow readers to reproduce the results in this paper are available at https://github.com/north-numerical-computing/integer-matrix-multiply-experiments and in the supplementary materials ( integer-matrix-multiply-experiments-main.zip [97.8KB]). [Formula: see text]

SIAM Journal on Scientific ComputingVol. 48(6)
University of Leeds (GB), University of Manchester (GB), University of Tennessee at Knoxville (US)
Nvidia, University of Oregon, Engineering and Physical Sciences Research Council
Openalex Percentile: Top 0%
Numerical Methods and Algorithms
27.90
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon