Graph Neural Network Training with Data Tiering

TL;DR

This paper introduces a data tiering approach for GNN training that reduces memory bottlenecks and communication overhead, significantly accelerating training on large-scale graphs.

Contribution

It presents a novel data tiering method leveraging graph structure and training insights to optimize data placement and access in multi-GPU GNN training.

Findings

CPU-GPU traffic reduced by 87-95%

Training speed improved by 1.6-2.1x

Effective on graphs with hundreds of millions of nodes

Abstract

Graph Neural Networks (GNNs) have shown success in learning from graph-structured data, with applications to fraud detection, recommendation, and knowledge graph reasoning. However, training GNN efficiently is challenging because: 1) GPU memory capacity is limited and can be insufficient for large datasets, and 2) the graph-based data structure causes irregular data access patterns. In this work, we provide a method to statistical analyze and identify more frequently accessed data ahead of GNN training. Our data tiering method not only utilizes the structure of input graph, but also an insight gained from actual GNN training process to achieve a higher prediction result. With our data tiering method, we additionally provide a new data placement and access strategy to further minimize the CPU-GPU communication overhead. We also take into account of multi-GPU GNN training as well and we demonstrate the effectiveness of our strategy in a multi-GPU system. The evaluation results show that our work reduces CPU-GPU traffic by 87-95% and improves the training speed of GNN over the existing solutions by 1.6-2.1x on graphs with hundreds of millions of nodes and billions of edges.