Welcome to Fangxin Liu’s Homepage~
Fangxin (Leon) Liu is an Assistant Professor and Ph.D. supervisor in the School of Computer Science at Shanghai Jiao Tong University (SJTU). I also serve as a Research Fellow at the Shanghai Qi Zhi Institute. My research interests include neural network acceleration (e.g., mixed-precision computing and SW/HW co-design), in-memory computing, and brain-inspired neuromorphic computing.
I received my Ph.D. degree from SJTU in 2023, advised by Prof. Li Jiang. To date, I have published over 60 papers as first or corresponding author, with more than 40 in CCF Tier A venues, including top-tier architecture conferences and journals such as ISCA, MICRO, ASPLOS, HPCA, PPoPP, DAC, IEEE TC, TPDS, and TCAD. My work has been recognized with honors such as the Spark Award from HUAWEI, the Outstanding Doctoral Dissertation Award from the Shanghai Computer Society, the ACM China Shanghai Excellent Doctoral Dissertation Award, and the Best Paper Award at DATE 2022.
My research has been successfully adopted by leading technology companies, including Huawei, Ant Group, ZTE, and Yizhu Tech., to advance real-world applications such as efficient LLM inference, low-precision deployment, and optimized AI compilation. For instance, my work on mixed-precision computing and SW/HW co-design has enabled up to 40% reductions in computational costs for large-scale AI deployments. I am committed to bridging cutting-edge research with practical industry solutions and welcome opportunities for further collaboration.
🔥 Recruitment
Our team is actively seeking self-motivated PhD, Master, and Undergraduate students interested in Computer Architecture, Efficient AI acceleration, and PIM Design. If you are interested, please email me your CV.
News
🕒 Click to view all Archived News (2025 - 2022)
Research
His research bridges the gap between Cutting-edge Architecture and Practical AI Solutions, focusing on:
- LLM & Generative AI Acceleration (Mixed-precision, Quantization, MoE)
- In-Memory Computing (RRAM/DRAM-based PIM/CiM)
- Brain-inspired Computing (Spiking Neural Networks, HDC)
- AI Compilation & HW/SW Co-design