LPS-GNN : Deploying Graph Neural Networks on Graphs with 100-Billion Edges

TL;DR

LPS-GNN is a scalable, efficient GNN framework capable of processing 100 billion edges on a single GPU in 10 hours, with significant improvements in prediction accuracy and deployment on Tencent.

Contribution

The paper introduces LPMetis, a superior graph partitioning algorithm, and a subgraph augmentation strategy, enhancing GNN scalability and performance on large-scale graphs.

Findings

LPS-GNN processes 100 billion edges on a single GPU in 10 hours.

Achieves 13.8% improvement in User Acquisition scenarios.

Outperforms SOTA models with 8.24% to 13.89% performance lift in online applications.

Abstract

Graph Neural Networks (GNNs) have emerged as powerful tools for various graph mining tasks, yet existing scalable solutions often struggle to balance execution efficiency with prediction accuracy. These difficulties stem from iterative message-passing techniques, which place significant computational demands and require extensive GPU memory, particularly when dealing with the neighbor explosion issue inherent in large-scale graphs. This paper introduces a scalable, low-cost, flexible, and efficient GNN framework called LPS-GNN, which can perform representation learning on 100 billion graphs with a single GPU in 10 hours and shows a 13.8% improvement in User Acquisition scenarios. We examine existing graph partitioning methods and design a superior graph partition algorithm named LPMetis. In particular, LPMetis outperforms current state-of-the-art (SOTA) approaches on various evaluation metrics. In addition, our paper proposes a subgraph augmentation strategy to enhance the model's predictive performance. It exhibits excellent compatibility, allowing the entire framework to accommodate various GNN algorithms. Successfully deployed on the Tencent platform, LPS-GNN has been tested on public and real-world datasets, achieving performance lifts of 8. 24% to 13. 89% over SOTA models in online applications.