PDGNN: Efficient Micro-batch GNN Training via Degree-Pruned Partitioning and Redundancy Elimination
ACM Transactions on Architecture and Code Optimization · 31.12.2025 · 22 (4) · 1-25
DOI: 10.1145/3767325 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Transactions on Architecture and Code Optimization, online erschienen am 12.12.2025. Verfasst von Lizhi Zhang und 7 weiteren Autor:innen. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: National Natural Science Foundation of China, National Key Research and Development Program of China, Open Fund of PDL.
Abstract (Originalsprache)
Graph Neural Networks (GNNs) have become essential for graph-based applications. Existing GNN training frameworks primarily adopt sampling-based mini-batch approach, which samples vertex neighborhoods to construct subgraphs for training iteratively, but faces a memory-accuracy tradeoff. To address this, naive micro-batch training with gradients accumulation can be applied, yet introduces severe redundancy from repeated neighbor sampling. Recent micro-batch training methods overcome this by adopting batch-level graph partitioning strategies that partition full-batch subgraph sampled from all target vertices to generate micro-batches. Nevertheless, these approaches struggle to balance partitioning overhead with redundancy. This article proposes PDGNN, a system designed for efficient micro-batch GNN training on single machine, reducing both training time and memory consumption. First, we propose a degree-pruning-based graph partitioning method that efficiently processes on simplified and sparsified multi-level bipartite graphs by removing high-degree vertices, significantly reducing partitioning overhead. Second, we propose a dependency-aware redundancy matching and elimination technique that maximizes inter-micro-batch redundancy matching, and eliminates redundant computation by embedding caching and reuse while maintaining model accuracy. Experiments on real-world datasets demonstrate that PDGNN substantially improves the efficiency of large-scale GNN training, achieving up to 3.5× speedup in epoch time, 46% lower peak GPU memory consumption, and 64% reduced peak CPU memory consumption compared to state-of-the-art batch-level training methods.
Förderung
- National Natural Science Foundation of China
- National Key Research and Development Program of China
- Open Fund of PDL
Zitieren
APA
Zhang, L., Jia, M., Gong, P., Lai, Z., Li, D., Fu, Y., Shen, A., & Lu, K. (2025). PDGNN: Efficient Micro-batch GNN Training via Degree-Pruned Partitioning and Redundancy Elimination. ACM Transactions on Architecture and Code Optimization, 22(4), 1-25. https://doi.org/10.1145/3767325
Vancouver
Zhang L, Jia M, Gong P, Lai Z, Li D, Fu Y, et al. PDGNN: Efficient Micro-batch GNN Training via Degree-Pruned Partitioning and Redundancy Elimination. ACM Transactions on Architecture and Code Optimization. 2025;22(4):1-25. doi:10.1145/3767325
BibTeX
@article{Zhang2025,
author = {Zhang, Lizhi and Jia, Menghan and Gong, Ping and Lai, Zhiquan and Li, Dongsheng and Fu, Yongquan and Shen, Ao and Lu, Kai},
title = {PDGNN: Efficient Micro-batch GNN Training via Degree-Pruned Partitioning and Redundancy Elimination},
journal = {ACM Transactions on Architecture and Code Optimization},
year = {2025},
volume = {22},
number = {4},
pages = {1--25},
doi = {10.1145/3767325}
}
EUVIMED