The goal of our research is to explore and build the next-generation hardware based on post-CMOS emerging devices, e.g. memristors. From top-down, we examine the existing algorithms and determine the most crucial task to accelerate with our hardware. Examples include but not limited to matrix multiplication in A.I./machine learning tasks. In the meanwhile, we optimize and characterize the device performance, including their nonlinear dynamics, and build novel circuit/architecture/algorithm from bottom up.
The followings are example topics. If you are interested in joining us (as research postgraduate (e.g. Ph.D., MPhil.,) students, research assistant, postdoc) and would like to know about the on-going topics, please feel free to contact me at canl@hku.hk.
1. CMOS compatible memristor device integration and nanofabrication
References:
-
[Nature Electronics], 2020
P. Lin, C. Li, et al., Three-dimensional memristor circuits as complex neural networks. Preview PDF -
[Nature Nanotechnology], 2019
S. Pi, C. Li, et al., Memristor crossbar arrays with 6-nm half-pitch and 2-nm critical dimension. Preview PDF -
[Nature Communications], 2017
C. Li, et al., Three-Dimensional Crossbar Arrays of Self-rectifying Si/SiO2/Si Memristors. Open Access PDF
2. Novel application, circuit, device, and their co-design.
References:
-
[Nature Communications], 2023
M. Jiang, K. Shan, C. He, C. Li, Efficient combinatorial optimization by quantum-inspired parallel annealing in analogue memristor crossbar. Open Access PDF -
[IEDM], 2022
M. Jiang, K. Shan,… C. Li, An efficient synchronous-updating memristor-based Ising solver for combinatorial optimization. PDF -
[Nature Communications], 2020
C. Li, et al., Analog content addressable memories with memristors. (Presented at HPE internal TechCon 2020 as “Analog content addressable memory for explainable and efficient machine learning”. (acceptance rate of 4.38%)) Open Access PDF -
[Nature Electronics], 2018
C. Li, et al., Analog signal and image processing with large memristor crossbars.Preview PDF -
[Nature Electronics], 2018
H. Jiang†, C. Li†, et al., Provable Key Destruction with Large Memristor Crossbar. Preview PDF
3. A.I./Machine learning hardware accelerators based on emerging devices.
References:
-
[Nature Communications], 2022 R. Mao, B. Wen, … C. Li Experimentally validated memristive memory augmented neural network with efficient hashing and similarity search. Open Access PDF
-
[Nature Communications], 2018
C. Li, et al., Efficient and self-adaptive in-situ learning in multilayer memristor neural networks. Open Access PDF -
[IMW], 2020
C. Li, et al., CMOS-integrated nanoscale memristive crossbars for CNN and optimization acceleration. PDF -
[Nature Machine Intelligence], 2019
C. Li, et al., Long short-term memory networks in memristor crossbars. Preview PDF -
[Nature Machine Intelligence], 2019
Z. Wang†, C. Li†, P. Lin†, et al., In situ training of feedforward and recurrent convolutional memristor networks. Preview PDF -
[Nature Electronics], 2019
Z. Wang†, C. Li†, et al., Reinforcement learning with analog memristor arrays. Preview PDF