Next Article in Journal
Skeletal Image Features Based Collaborative Teleoperation Control of the Double Robotic Manipulators
Previous Article in Journal
Application of FPGA Devices in Network Security: A Survey
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

Automatic Sparse Matrix Format Selection via Dynamic Labeling and Clustering on Heterogeneous CPU–GPU Systems

School of Microelectronics, South China University of Technology, Guangzhou 510006, China
*
Author to whom correspondence should be addressed.
Electronics 2025, 14(19), 3895; https://doi.org/10.3390/electronics14193895
Submission received: 4 September 2025 / Revised: 28 September 2025 / Accepted: 28 September 2025 / Published: 30 September 2025

Abstract

Sparse matrix–vector multiplication (SpMV) is a fundamental kernel in high-performance computing (HPC) whose efficiency depends heavily on the storage format across central processing unit (CPU) and graphics processing unit (GPU) platforms. Conventional supervised approaches often use execution time as training labels, but our experiments on 1786 matrices reveal two issues: labels are unstable across runs due to execution-time variability, and single-label assignment overlooks cases where multiple formats perform similarly well. We propose a dynamic labeling strategy that assigns a single label when the fastest format shows clear superiority, and multiple labels when performance differences are small, thereby reducing label noise. We further extend feature analysis to multi-dimensional structural descriptors and apply clustering to refine label distributions and enhance prediction robustness. Experiments demonstrate 99.2% accuracy in hardware (CPU/GPU) selection and up to 98.95% accuracy in format prediction, with up to 10% robustness gains over traditional methods. Under cost-aware, end-to-end evaluation that accounts for feature extraction, prediction, conversion, and kernel execution, CPUs achieve speedups up to 3.15× and GPUs up to 1.94× over a CSR baseline. Cross-round evaluations confirm stability and generalization, providing a reliable path toward automated, cross-platform SpMV optimization.
Keywords: sparse matrix–vector multiplication; dynamic labeling; clustering; format prediction; heterogeneous computing sparse matrix–vector multiplication; dynamic labeling; clustering; format prediction; heterogeneous computing

Share and Cite

MDPI and ACS Style

Shi, Z.; Zou, Y.; Song, X. Automatic Sparse Matrix Format Selection via Dynamic Labeling and Clustering on Heterogeneous CPU–GPU Systems. Electronics 2025, 14, 3895. https://doi.org/10.3390/electronics14193895

AMA Style

Shi Z, Zou Y, Song X. Automatic Sparse Matrix Format Selection via Dynamic Labeling and Clustering on Heterogeneous CPU–GPU Systems. Electronics. 2025; 14(19):3895. https://doi.org/10.3390/electronics14193895

Chicago/Turabian Style

Shi, Zheng, Yi Zou, and Xianfeng Song. 2025. "Automatic Sparse Matrix Format Selection via Dynamic Labeling and Clustering on Heterogeneous CPU–GPU Systems" Electronics 14, no. 19: 3895. https://doi.org/10.3390/electronics14193895

APA Style

Shi, Z., Zou, Y., & Song, X. (2025). Automatic Sparse Matrix Format Selection via Dynamic Labeling and Clustering on Heterogeneous CPU–GPU Systems. Electronics, 14(19), 3895. https://doi.org/10.3390/electronics14193895

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop