报告人：   胡国华，香港中文大学 助理教授

报告地点：  新校区物理楼345

报告时间： 2024年4月9日（周二）下午4:30-5:30

报告题目： Implementation of neuromorphic computing using solution-processed low-dimensional materials

个人简介：

Dr Guohua Hu is an Assistant Professor at the Electronic Engineering Department, The Chinese University of Hong Kong. He joined the Department in 2019. The research focus of his group is solution-phase processing of low-dimensional materials and exploring their broad applications in printed electronics, including thin-film transistors and memristive electronics. Before joining CUHK, he was a postdoctoral research associate at the Cambridge Graphene Centre. He earned his BEng and PhD degrees from the University of Electronic Science and Technology of China and the University of Cambridge in 2013 and 2018, respectively.

报告简介：

With the rapid advancement of artificial intelligence and machine learning, the traditional von Neumann computing is facing challenges in the computational power and energy consumption. Inspired by the human brain, neuromorphic computing by mapping the structural and functional architectures of the neural networks has emerged as a promising alternative paradigm. Low-dimensional materials, with their unique electronic properties, allow device fabrication and engineering towards the implementation of neuromorphic computing. In this talk, I will discuss the recent progress of our research on neuromorphic computing using solution-processed low-dimensional materials and devices. The first part of my talk will focus on solution processing of low-dimensional materials, such as two-dimensional materials and carbon nanotubes, for the development of printed electronics. Specifically, I will discuss the fabrication of memristor and memristive transistor devices, and the exploitation of the device characteristics in the design of artificial neurons and synapses. On this basis, the second part of my talk will discuss the implementation of neuromorphic computing approaches, including the convolutional computing, spiking neuromorphic computing, and reservoir computing, using the devices and the artificial neurons and synapses. The demonstrations leveraging the device characteristics hold the promise to enable efficient computation in autonomous driving, virtual reality, medical diagnosis, industrial automation, and beyond.