I am a full professor at the Institute of Computing Technology (ICT), Chinese Academy of Sciences (CAS). My research interests generally lie in a broad range of computer architecture and artificial intelligence. In addition to designing efficient architecture, compilers, and systems for accelerating artificial intelligence workloads, I am also interested in applying artificial intelligence techniques to ease the design/modeling/optimization of circuits, architecture, compilers, and systems. I have published about 100 papers on both top-tier computer architecture venues including ISCA/MICRO/ASPLOS/HPCA and artificial intelligence venues including IJCAI/AAAI/ICML/NeurIPS/ICLR. I am also on the editorial board of the IEEE Transactions on Computers and Pattern Recognition.

• Sys4ML: Architecture/compilation/systems for machine learning
• ML4Sys: Machine learning for circuit/architecture/compiler/system design
• Other topics of computer architecture and artificial intelligence

• 2019 - Now: Professor, Institute of Computing Technology, CAS.
• 2015 - 2019: Associate Professor, Institute of Computing Technology, CAS.
• 2014 - 2015: Postdoctoral Researcher, Carnegie Mellon University.
• 2012 - 2014: Researcher, IBM Research China.
• 2007 - 2012: Ph.D. in Computer Science, Institute of Computing Technology, CAS.
• 2002 - 2007: B.S. in Computer Science, Tongji University.

• Sys4ML: Architecture/compilation/systems for machine learning
  1. [ISCA'24] Camrbicon-D: full-network differential acceleration for diffusion models. [PDF]
  2. [MICRO'23] Cambricon-U: A systolic random increment memory architecture for unary computing. [PDF]
  3. [MICRO'23] Cambricon-R: A fully fused accelerator for real-time learning of neural scene representation. [PDF]
  4. [ASPLOS'23] Heron: Automatically constrained high-performance library generation for deep learning accelerators. [PDF]
  5. [MICRO'22] Cambricon-P: A bitflow architecture for arbitrary precision computing. [PDF]
• ML4Sys: Machine learning for architecture/compilation/system design
  1. [DAC'24] Revisiting automatic pipelining: gate-level forwarding and speculation. [PDF]
  2. [IJCAI'24] Automated CPU design by learning from input-output examples. [PDF]
  3. [ICML'24] AutoOS: Make your OS more powerful by exploiting large language models. [PDF]
  4. [ICLR'23] Efficient tensor program optimization with diversity-based active learning. [PDF]
  5. [ICML'22] BabelTower: Learning to auto-parellelized program translatio. [PDF]

• 2022, Young Scientist Award of the Chinese Academy of Sciences
• 2022, MICRO'22 Best Paper Runner-Up
• 2015, CGO'15 Best Paper Nominee
• 2015, HPEC'15 Best Paper Nominee
• 2012, IISWC'12 Best Paper Nominee