Graph

Graph Neural Network (GNN) has emerged as an important workload for learning on graphs. With the size of graph data and the complexity of GNN model architectures increasing, developing an efficient GNN system grows more important.