Accelerating GNN Training by Adapting Large Graphs to Distributed Heterogeneous Architectures | IEEE Journals & Magazine | IEEE Xplore