SUGAR: Efficient Subgraph-level Training via Resource-aware Graph Partitioning

TL;DR

SUGAR introduces a resource-aware graph partitioning method that enables efficient subgraph-level training of GNNs, significantly reducing runtime and memory usage for large-scale graphs, thus facilitating IoT deployment.

Contribution

It proposes a novel resource-aware graph partitioning approach for GNN training, enabling efficient subgraph-level learning with theoretical analysis and extensive experimental validation.

Findings

Achieves up to 33x runtime speedup

Reduces memory usage by 3.8x

Effective on large-scale graph benchmarks

Abstract

Graph Neural Networks (GNNs) have demonstrated a great potential in a variety of graph-based applications, such as recommender systems, drug discovery, and object recognition. Nevertheless, resource-efficient GNN learning is a rarely explored topic despite its many benefits for edge computing and Internet of Things (IoT) applications. To improve this state of affairs, this work proposes efficient subgraph-level training via resource-aware graph partitioning (SUGAR). SUGAR first partitions the initial graph into a set of disjoint subgraphs and then performs local training at the subgraph-level. We provide a theoretical analysis and conduct extensive experiments on five graph benchmarks to verify its efficacy in practice. Our results show that SUGAR can achieve up to 33 times runtime speedup and 3.8 times memory reduction on large-scale graphs. We believe SUGAR opens a new research direction towards developing GNN methods that are resource-efficient, hence suitable for IoT deployment.