DiskGNN: Bridging I/O Efficiency and Model Accuracy for Out-of-Core GNN Training

TL;DR

DiskGNN is a system that efficiently trains large-scale GNNs out-of-core by combining offline sampling, optimized data packing, and pipelined training to achieve high I/O efficiency without sacrificing accuracy.

Contribution

DiskGNN introduces offline sampling and optimized data management techniques to improve I/O efficiency and maintain model accuracy in out-of-core GNN training.

Findings

Over 8x speedup compared to state-of-the-art systems

Maintains the same model accuracy as in-memory training

Effective use of offline sampling and feature packing techniques

Abstract

Graph neural networks (GNNs) are machine learning models specialized for graph data and widely used in many applications. To train GNNs on large graphs that exceed CPU memory, several systems store data on disk and conduct out-of-core processing. However, these systems suffer from either read amplification when reading node features that are usually smaller than a disk page or degraded model accuracy by treating the graph as disconnected partitions. To close this gap, we build a system called DiskGNN, which achieves high I/O efficiency and thus fast training without hurting model accuracy. The key technique used by DiskGNN is offline sampling, which helps decouple graph sampling from model computation. In particular, by conducting graph sampling beforehand, DiskGNN acquires the node features that will be accessed by model computation, and such information is utilized to pack the target node features contiguously on disk to avoid read amplification. Besides, \name{} also adopts designs including four-level feature store to fully utilize the memory hierarchy to cache node features and reduce disk access, batched packing to accelerate the feature packing process, and pipelined training to overlap disk access with other operations. We compare DiskGNN with Ginex and MariusGNN, which are state-of-the-art systems for out-of-core GNN training. The results show that DiskGNN can speed up the baselines by over 8x while matching their best model accuracy.