Jin-Liang Shi

UltraGNN: A Sparse-Operator-Aware Framework for Accelerating Graph Neural Networks on Tensor Cores

Graph Neural Networks (GNNs) have achieved widespread success from social networks to AI-for-Science. Most existing GNN frameworks adopt scatter-first (edge-centric) or gather-first (vertex-centric) scheduling paradigms for message passing. However, these paradigms are closely tied to traditional CUDA-core execution mo...