[Blog] Large-Scale Training Of Graph Neural Networks

Many graph applications deal with giant scale. Social networks, recommendation and knowledge graphs have nodes and edges in the order of hundreds of millions or even billions of nodes. For example, a recent snapshot of the friendship network of Facebook contains 800 million nodes and over 100 billion links.

This is a companion discussion topic for the original entry at https://www.dgl.ai/blog/2019/06/13/giant.html

Thanks for your excellent work.
And I have a question. It seems that DGL stores a graph on one machine, but uses some samplers on different machines to sample it. I don’t understand why it can improve performance since it just requires more network communications.