Short Bio

Hemeng Wang is a third-year PhD student at China University of Petroleum-Beijing, under the supervision of Prof. Weifeng Liu. He earned his Bachelor's degree in Computer Science and Technology from China University of Petroleum-Beijing in 2021. Hemeng's research interests include high-performance computing, parallel computing, and matrix computation. Currently, his research focuses on the memory access and communication of matrix multiplication, and he has proposed KAMI, a novel method to accelerate general matrix-matrix multiplication (GEMM). He is a Student Member of IEEE and CCF.

Publications

• [SC '25] Hemeng Wang, Yang Du, Sidu Li, Qingxiao Sun, Weifeng Liu. KAMI: Communication-Avoiding General Matrix Multiplication within a Single GPU. 38th International Conference for High Performance Computing, Networking, Storage, and Analysis. 2025. [PDF] [Slides] [DOI] [Bibtex] [Code]
• [CCF THPC] Hemeng Wang, Wenqing Lin, Qingxiao Sun, Weifeng Liu. νGNN: Non-Uniformly partitioned full-graph GNN training on mixed GPUs. CCF Transactions on High Performance Computing. 2025. [PDF] [DOI] [Bibtex] [Code]
  
• [CLUSTER '24] Wenqing Lin, Hemeng Wang, Haodong Deng, Qingxiao Sun. ScalFrag: Efficient Tiled-MTTKRP with Adaptive Launching on GPUs. 26th IEEE International Conference on Cluster Computing. 2024. [PDF] [Slides] [DOI] [Bibtex] [Code]