Associate Professor
Institute of Computing Technology
Chinese Academy of Sciences

I am an associate professor at Insititute of Computing Technology, Chinese Academy of Sciences. I received Ph.D degree in School of Computer Science and Engineering at Beihang University, under the supervision of Prof. Yi Liu and Prof. Hailong Yang. As a member of the High Performance Computer Research Center, led by Prof. Guangming Tan, my research interests include high-performance computing, AI system, and AI4S (AI for Science) system. My recent reasearch focus on hybrid parallelism, elastic scaling, compilation optimization for efficient AI/AI4S model training and inference. I received CCF Outstanding Doctoral Dissertation Award in High-Performance Computing, and ACM China Doctoral Dissertation Award in SIGHPC.


🎤 Talks

  • Oct. 2026 — Load Balancing Methods for Long-Context LLM Training. HPCChina 2026, Shenzhen, China(长上下文训练的负载均衡方法,高性能计算年会,深圳).
  • Sep. 2026 — Parallel Training Framework for Large MLIP Models. Supercomputing Innovation Alliance Conference, Yining, Xinjiang, China(大原子模型并行训练框架,超级计算创新联盟会议,新疆伊宁).

📑 Publications

💻 Projects

  • LunaEquivariance — A high-performance, open-source library of SO(3)-equivariant operators for graph neural networks and machine learning interatomic potentials (MLIPs). It provides optimized tensor products, fused graph convolutions, and attention, with PyTorch integration, FP32/FP64 support, and derivatives up to third order.
    Paper: LunaTP (arXiv) · GitHub Repository

🏛️ Fundings

  • China National Postdoctoral Program for Innovative Talents (BX Program)
  • National Natural Science Foundation of China (NSFC) Young Scientists Fund
  • Beijing Natural Science Foundation Young Scientists Program
  • Chinese Academy of Sciences Special Research Assistant Project