EUVIMEDEUVIMED entdecken
Crossref

Efficient Graph Data Access for Out-of-Memory GPU Streaming Graph Processing

Qiange Wang, Yongze Yan, Hongshi Tan, Cheng Chen, Cheng Zhao, Jiaming Tian, Jiaxin Jiang, Xiaoliang Cong, Yanfeng Zhang, Ge Yu, Weng-Fai Wong, Bingsheng He

Proceedings of the VLDB Endowment · 2025-07 · 18 (11) · 3854-3867

DOI: 10.14778/3749646.3749659 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Proceedings of the VLDB Endowment, online erschienen am 04.09.2025. Verfasst von Qiange Wang und 11 weiteren Autor:innen. Die Arbeit wurde bisher 5-mal zitiert (Crossref) und nennt 54 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Leveraging GPUs' high parallelism can significantly improve the real-time computation efficiency of streaming graph processing. However, when a large-scale graph exceeds GPU memory capacity, CPU-GPU cooperative processing often results in substantial and irregular CPU-to-GPU data transfer overhead. This stems from the extensive redundant graph accesses during continuous computation, which can hardly be addressed by existing solutions. In this work, we present Grapin, an out-of-memory GPU streaming graph processing system designed to minimize graph data transfer via two effective techniques for eliminating redundant accesses: (1) Extending advanced incremental processing algorithms to GPUs by converting their heavyweight data dependency processing into GPU-friendly forms, eliminating redundant graph accesses from the computation side; and (2) providing a lightweight yet efficient GPU hot subgraph management framework that finely caches the frequently accessed dynamic subgraphs in a vertex-centric manner. Experimental results demonstrate that Grapin can efficiently process large-scale streaming graphs with billions of edges on a single NVIDIA A5000 GPU. Enabling incremental computation reduces data transfer by 61%, and the integration of GPU hot subgraph reuse further reduces the remaining transfer by 72%, resulting in a total reduction of 89%. Compared with CPU-based solutions, Grapin achieves speedups ranging from 1.8x to 96.9x (17.9x on average).

Zitieren

APA

Wang, Q., Yan, Y., Tan, H., Chen, C., Zhao, C., Tian, J., Jiang, J., Cong, X., Zhang, Y., Yu, G., Wong, W., & He, B. (2025). Efficient Graph Data Access for Out-of-Memory GPU Streaming Graph Processing. Proceedings of the VLDB Endowment, 18(11), 3854-3867. https://doi.org/10.14778/3749646.3749659

Vancouver

Wang Q, Yan Y, Tan H, Chen C, Zhao C, Tian J, et al. Efficient Graph Data Access for Out-of-Memory GPU Streaming Graph Processing. Proceedings of the VLDB Endowment. 2025;18(11):3854-3867. doi:10.14778/3749646.3749659

BibTeX
@article{Wang2025,
  author = {Wang, Qiange and Yan, Yongze and Tan, Hongshi and Chen, Cheng and Zhao, Cheng and Tian, Jiaming and Jiang, Jiaxin and Cong, Xiaoliang and Zhang, Yanfeng and Yu, Ge and Wong, Weng-Fai and He, Bingsheng},
  title = {Efficient Graph Data Access for Out-of-Memory GPU Streaming Graph Processing},
  journal = {Proceedings of the VLDB Endowment},
  year = {2025},
  volume = {18},
  number = {11},
  pages = {3854--3867},
  doi = {10.14778/3749646.3749659}
}
Kostenlos starten →