We present BatchGNN, a distributed CPU system that showcases techniques that
can be used to efficiently train GNNs on terabyte-sized graphs. It reduces
communication overhead with macrobatching in which multiple minibatches'
subgraph sampling and feature fetching are batched into one communication relay
to reduce redundant feature fetches when input features are static. BatchGNN
provides integrated graph partitioning and native GNN layer implementations to
improve runtime, and it can cache aggregated input features to further reduce
sampling overhead. BatchGNN achieves an average $3\times$ speedup over DistDGL
on three GNN models trained on OGBN graphs, outperforms the runtimes reported
by distributed GPU systems $P^3$ and DistDGLv2, and scales to a terabyte-sized
graph.

BatchGNN 是一个分布式 CPU 系统，可用于高效地在千兆字节级别的图形上训练图神经网络，通过在一个通信中绑定多个子图采样和特征提取以减少冗余特征提取、提供整合图分割和本地 GNN 层实现来提高运行时间，可以缓存聚合输入特征以进一步减少采样开销，相对于 DistDGL，GNN 的平均速度加快了 3 倍，在三个 OGBN 图表上训练的 GNN 模型优于分布式 GPU 系统 P3 和 DistDGLv2 的运行时间，并且可以扩展到千兆字节级别的图形。