High-Performance Mixed-Precision Matrix Multiplication via Tile-Centric Design on Modern Architectures
Name
42979_2025_4575_ReferencePDF.pdf
Size
5.73 MB
Format
Adobe PDF
Checksum (MD5)
9bea7906b13124eeacb2421cbfc7552c
Author(s)
Zhang, Qiao
Alomairy, Rabab
Wang, Dali
Gu, Zhuowei
Cao, Qinglei
Date Issued
December 20, 2025
Journal
SN Computer Science
Publisher
Springer Nature Singapore
Citation
Zhang, Q., Alomairy, R., Wang, D. et al. High-Performance Mixed-Precision Matrix Multiplication via Tile-Centric Design on Modern Architectures. SN COMPUT. SCI. 7, 24 (2026).
Version
Author's final manuscript
Abstract
General Matrix Multiplication (GEMM) is a critical operation underpinning a wide range of applications in high-performance computing (HPC) and artificial intelligence (AI). The emergence of hardware optimized for low-precision arithmetic necessitates a reevaluation of numerical algorithms to leverage mixed-precision computations, achieving improved performance and energy efficiency. This research presents an adaptive mixed-precision GEMM framework that enables support for various precision formats at fine-grained tile and block levels, offering a reliable foundation for trustworthy mixed-precision computations. Furthermore, we leverage the PaRSEC runtime system to effectively balance workloads across diverse architectures. The performance exhibits strong scalability across both homogeneous platforms (Intel CPU-based systems and the ARM CPU-based Fugaku supercomputer) and heterogeneous systems (Nvidia V100, A100, and H100 GPU-based platforms, as well as the AMD GPU-based Frontier supercomputer). This work aims to improve computational efficiency and accuracy by bridging algorithmic innovations with hardware capabilities, fostering transformative advancements across a wide range of applications.
MIT Department
Massachusetts Institute of Technology. Computer Science and Artificial Intelligence Laboratory
Terms of Use
Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.
Persistent DSpace Link
DOI of Published Version
https://doi.org/10.1007/s42979-025-04575-0
