Fetching the paper…

BGL: GPU-Efficient GNN Training by Optimizing Graph Data I/O and Preprocessing · Around