Fast Sparse GPU Kernels for Accelerated Training of Graph Neural Networks

This work proposes fast sparse GPU kernels tailored for training graph neural networks (GNNs). By carefully optimizing memory access patterns, load balancing, and sparse computation primitives, our kernels significantly reduce training time for a variety of GNN architectures and large-scale graph datasets.

Recommended citation: **Ruibo Fan**, Wei Wang, and Xiaowen Chu, "Fast Sparse GPU Kernels for Accelerated Training of Graph Neural Networks," in *Proceedings of the IEEE International Parallel and Distributed Processing Symposium (IPDPS)*, 2023.
Download Paper | Code | Download Bibtex