Welcome to the Cognitive and Adaptive Neuromorphic computing Lab (CANLab) at the University of Hong Kong! Our research develops analog computing hardware from emerging memory devices, spanning micro/nanofabrication, prototype chip design and measurement, and hardware-algorithm co-design. The work connects device physics with reliable in-memory computing for machine learning, search, genomic analysis, and combinatorial optimization. Please refer to the Research page for more details.

Recent Research Highlights

• Fault-free analog computing with imperfect hardware, Nature Electronics (2026).
• Trustworthy tree-based machine learning with MoS2 flash-based analog content-addressable memory, Nature Communications (2026).
• Hardware-adaptive associative memory on memristor crossbars, Nature Communications (2026).
• Sb-contacted MoS2 flash memory for in-memory searches, Nature Nanotechnology (2026).

See Publications for the publication list and News for recent coverage.

We are actively recruiting Postdoctoral Fellows, in particular for candidates with expertise in:

• HfO2-based FeRAM, FeFET, and 2T1C/1T1C device, and BEOL-compatible integration / 3D stacking.
• FPGA/PCB-based system building and experimental benchmarking of prototype chips;
• Hardware-software codesign for Transformer/LLM algorithms.

Interested candidates please send your CV to Prof. Can Li at canl@hku.hk.

This site is powered by Jekyll and hosted on GitHub Pages. Any opinions on this site are mine alone.