Join Zoom Meeting  

https://hku.zoom.us/j/93981531659?pwd=O54cidbwwfblgJ19EHdgDN0GwX1CnP.1

Meeting ID: 939 8153 1659

Password: 624474

Abstract

With hundreds of zettabytes of data generated annually, artificial intelligence (AI) demands ever-increasing computing performance. However, data transfer both between and within chips remains a critical bottleneck for system performance scaling. Enhancing data transfer bandwidth through monolithic three-dimensional (M3D) integration and alleviating data transfer loads via brain-inspired computing paradigms are promising solutions. In this talk, I will present my recent work on back-end-of-line (BEOL)-compatible oxide electronics, which play a pivotal role in these solutions: (1) To tackle the critical challenge of reliability in amorphous oxide semiconductor (AOS) transistors, a prevalent BEOL-compatible active device option in M3D integrated circuits (ICs), I have developed a cost-effective fluorination-in-packaging technique that could efficiently repair oxygen defects at a low thermal budget. (2) Recognizing the inefficiency of off-chip voltage conversion in 3D-ICs, I have developed BEOL-compatible AOS power transistor technology and experimentally demonstrated switched-capacitors DC-DC converters for efficient on-chip voltage conversion in heterogenous 3D integrated systems. (3) To yield more power-performance-area-cost advantages in brain-inspired computing hardware, I have developed a BEOL-compatible selective-area doping technique with an extremely high concentration in strongly correlated oxides, enabling the monolithic integration of VO2-based artificial neurons and synapses for efficient homotypic spiking neural platforms. These works underscore the significance of BEOL-compatible oxide electronics in advancing M3D integration and brain-inspired computing technologies, shaping the future of integrated circuits in the era of AI.

Biography

Dr. Sunbin Deng is currently a postdoctoral fellow working with Prof. Suman Datta at Georgia Institute of Technology, USA. His research interests include back-end-of-line (BEOL)-compatible nanoelectronics and related processing technologies for heterogenous/monolithic 3D integration, alternative computing, and information displays. Dr. Deng received his B.Sc. degree in Optical Information Science and Technology from Huazhong University of Science and Technology (HUST) in 2014 and Ph.D. degree in Electronic and Computer Engineering (under the supervision of Prof. Hoi-Sing Kwok) from the Hong Kong University of Science and Technology (HKUST) in 2020. From 2021 to 2022, he was a postdoctoral researcher at Purdue University, West Lafayette, USA, where he studied strongly correlated oxide electronics for neuromorphic and probabilistic computing. To date, Dr. Deng has co-authored over 80 papers in esteemed peer-reviewed journals (such as Sci. Adv., Nat. Commun., Adv. Mater., IEEE EDL, and IEEE T-ED) and top-tier global conferences (such as IEDM and VLSI), along with 1 book chapter and 6 patents. He has received several academic awards, including the Distinguished Paper Award at SID Display Week Symposium 2021 and the Young Leader Award from SID China.

ALL INTERESTED ARE WELCOME

For further information, please contact Prof. Mingxin Huang