I'm a Research Scientist at Compute Technology Lab, Alibaba DAMO Academy, Sunnyvale, since 2019/06.

Postdoc (2018-2019) and PhD (2014-2018) from UCSB under the supervision of Dr. Yuan Xie. M.S./B.S. from Tsinghua University.

I am a computer architect working on memory system and domain-specific (AI in particular) architecture. At DAMO Academy, we propose hardware innovations to address deep learning application challenges in Alibaba Cloud. Our on-going projects target at improving foundation model training efficiency and reducing LLM serving cost.

Before that, I worked on cloud FPGA solution to improve Alibaba's large scale GNN application (lead architect, FPGA prototype on ISCA-2022). Also, a 3D hybrid bonding near-data computing testchip to speedup Alibaba's recommandation system (testchip in ISSCC-2022).

Previously at UCSB, I worked on NVM and in-memory computing (PRIME, Pinatubo paper, 1500+ citations).
Please check my google scholar.

I am very honored to join Micro Hall of Fame in 2023.

Two GNN paper to appear on Micro-23, from our intern and Tsinghua collaborators.

PRIME paper is selected for inclusion in ISCA@50 25-year Retrospective 1996-2020.