Out-of-Memory Graph Processing Acceleration via Algorithmic-Hardware Codesign on FPGAs
IEEE Transactions on Computers · 2026-09 · 75 (9) · 3197-3211
DOI: 10.1109/tc.2026.3701992 ↗Kurz gefasst
Fachartikel in der Zeitschrift IEEE Transactions on Computers, 2026-09. Verfasst von Qianyu Cheng und 8 weiteren Autor:innen. Die Arbeit nennt 40 Literaturangaben. Gefördert von: Strategic Priority Research Program of the Chinese Academy of Sciences, National Natural Science Foundation of China, Natural Science Foundation of Jiangsu Province.
Abstract (Originalsprache)
Emerging FPGA devices are widely used in graph processing and achieve high performance by exploiting fine-grained parallelism and high-bandwidth memory (HBM) sub-systems with dozens of channels. However, the graph size increases rapidly and often exceeds the capacity of the FPGA device or one of the memory channels, incurring out-of-memory (OoM) faults and performance degradation. Existing approaches focus on improving the performance of on-device full graph processing, ignoring the support for hyper-scale graphs and cross-iteration acceleration. In this work, we propose SoGraph, a graph processing architecture for large graph acceleration on HBM-equipped FPGAs, cooperating with the host. To minimize PCIe traffic during the processing, we adopt three hardware-oriented optimizations: 1) activeness-driven subgraph scheduler, which minimizes on-device graph resident set; 2) lightweight state-aware processing extension in edge-centric paradigm; 3) bit-level update analyzer for to-be-transfer subgraph reduction. The results show that the prototype on Alveo U280 FPGA achieves an average of 22.3x performance speedup over the modified state-of-the-art FPGA design and 1.3x device energy efficiency over the GPU solution.
Förderung
- Strategic Priority Research Program of the Chinese Academy of Sciences
- National Natural Science Foundation of China
- Natural Science Foundation of Jiangsu Province
- Youth Innovation Promotion Association of the Chinese Academy of Sciences
- USTC Research Funds of the Double First-Class Initiative
- Students’ Innovation and Entrepreneurship Foundation of USTC
Zitieren
APA
Cheng, Q., Wu, J., Zheng, Z., Shi, C., Wang, T., Gong, L., Wang, C., Chen, H., & Xuehai, Z. (2026). Out-of-Memory Graph Processing Acceleration via Algorithmic-Hardware Codesign on FPGAs. IEEE Transactions on Computers, 75(9), 3197-3211. https://doi.org/10.1109/tc.2026.3701992
Vancouver
Cheng Q, Wu J, Zheng Z, Shi C, Wang T, Gong L, et al. Out-of-Memory Graph Processing Acceleration via Algorithmic-Hardware Codesign on FPGAs. IEEE Transactions on Computers. 2026;75(9):3197-3211. doi:10.1109/tc.2026.3701992
BibTeX
@article{Cheng2026,
author = {Cheng, Qianyu and Wu, Jing and Zheng, Zhendong and Shi, Chao and Wang, Teng and Gong, Lei and Wang, Chao and Chen, Huapin and Xuehai, Zhou},
title = {Out-of-Memory Graph Processing Acceleration via Algorithmic-Hardware Codesign on FPGAs},
journal = {IEEE Transactions on Computers},
year = {2026},
volume = {75},
number = {9},
pages = {3197--3211},
doi = {10.1109/tc.2026.3701992}
}
EUVIMED