题目：Memory and Computing Systems Based on Memristor Arrays
报告人：Professor Wei Lu （University of Michigan, IEEE Fellow）
摘要：Exponential growth of the semiconductor industry, historically driven by transistor scaling, is now facing fundamental challenges. In this talk, I will discuss an emerging class of devices that, by merging electronic with ionics, offer the potential to control the materials’ electronic properties in-situ and have led to promising memory and circuit concepts. These devices exhibit history-dependent, resistive switching behavior and are termed memristors (memory + resistor) or resistive memory (RRAM) devices. I will discuss our efforts on the development and optimization of memristor devices and integrated systems, including techniques of controlling the dynamic ionic migration processes and associated modeling efforts. Functional high-density crossbar arrays have been integrated directly on top of CMOS circuits using a back-end-of-line (BEOL) process, enabling hybrid non-volatile memory and reconfigurable circuit applications. Properly tuned devices also exhibit incremental conductance changes that are analogous to the behaviors of biological synapses and are well suited for hardware-based, bio-inspired neuromorphic logic systems. Prototype neuromorphic circuits based on memristor arrays have been shown to be able to perform tasks such as pattern recognition and image analysis in an unsupervised fashion for intelligent sensing, analysis and other data-intensive applications.
卢伟教授是电气电子工程师学会会士(IEEE Fellow)并长期担任国际半导体发展路线图(ITRS)的成员。他是非易失性存储器，纳米器件，异构整合芯片，及神经形态计算（类脑计算）电路的专家。他的成就包括阻变存储器（RRAM）的发展，RRAM Crossbar 阵列与CMOS电路的集成，神经形态计算芯片，纳米线晶体管及量子点电路。他还首先证明阻变器件可以有效地在神经形态系统应用领域模拟突触功能并实现类脑计算功能。他获得过NSF CAREER奖，Rexford E. Hall 发明奖, David E. Liddle 杰出研究奖, EECS 杰出研究成就奖。他发表了超过100篇期刊论文及拥有15项美国专利，至今被同行研究人员引用超过16,000次，h-因子为55。
Copyright © 2018 同济大学数学科学学院 版权所有.