Distinguished Speaker Series – “Memristive Dynamics Based Hardware Primitives for Efficient Computing”
Date:
27 Aug, 2021 (Fri)
Time:
3:00 pm - 4:30 pm
Webinar Link:
https://cutt.ly/HQuTGOf

Add to your calendar: iCal, Google Calendar

HKU Campus Map


Speaker

Prof. Yuchao Yang
Director
Center for Brain Inspired Chips,
Peking University

Abstract

Non-von Neumann architectures based on emerging memristive devices, such as ReRAM, PCM, etc., hold great prospect in constructing AI devices with extremely high energy efficiency that could bloom applications in edge computing. The introduction of analog nonvolatile memories allows both inference with high energy efficiency, and training as well. Throughput and energy efficiency of inference are strongly dependent on exercised neural network topologies, computational precision and network sparsity, and the chip design requires optimized mixed-signal circuit design, chip architecture and data flow, etc. On-chip local training is highly desirable for the application of deep neural networks in environment-adaptive edge platforms, which however is hindered by the high time and energy costs of training. Direct feedback alignment provides a viable choice for local learning when combined with transfer learning in convolutional layers, which can be achieved by exploiting the inherent stochasticity in the conductance states of phase change memory chip. Besides, a hardware spiking neural network circuit is constructed based on hybrid weight elements combing different memristive dynamics, which demonstrates temporal correlation detection after online unsupervised learning. We also developed a memristive optimizer hardware based on Hopfield network, which introduces transient chaos to simulate annealing in aid of jumping out of the local optima while ensuring convergence. A continuous function optimization problem as well as a NP-hard combinatorial optimization problem are experimentally demonstrated, therefore indicating great potential of the memristive optimizer as future computing platform for solving optimization problems in general.

 

Biography of the Speaker:

Yuchao Yang serves as Director of Center for Brain Inspired Chips at Peking University and Executive Director of Center for Brain Inspired Intelligence at Chinese Institute for Brain Science. His research interests include memristors, neuromorphic computing, and in-memory computing. He has published over 100 papers in high-profile journals and conferences such as Nature Electronics, Nature Communications, Nature Nanotechnology, Science Advances, Advanced Materials, Nano Letters, IEDM, etc. as well as 5 book chapters. His papers have been cited >5500 times, with an H-index of 33. He was invited to give >30 keynote/invited talks on international conferences and serves as TPC chair or member for 9 international conferences. Yuchao Yang serves as the Associate Editor for Nano Select and editorial board member of Chip, Scientific Reports and Science China Information Sciences. He was invited to guest edit 3 special issues and write 12 News & Views, review articles, etc. He is a member of IEEE, MRS and RSC. He is a recipient of the National Outstanding Youth Science Fund, Qiu Shi Outstanding Young Scholar Award, Wiley Young Researcher Award, MIT Technology Review Innovators Under 35 in China, and the EXPLORER PRIZE.


Organizer

Prof. K.K.Y. Wong

Most seminars are open to the general public, free of charge, unless otherwise stated. Registration is not required. Arrangement for car parking facilities on campus please contact us for details.

For enquiries, please contact:
Department of Electrical and Electronic Engineering,
Room 601, Chow Yei Ching Building,
Pokfulam Road, Hong Kong
Tel: (852) 3917 7093
Email: seminar@eee.hku.hk