Amorphous Oxide Semiconductor Memristors: Brain-inspired Computation

Royal Society of Chemistry eBooks(2023)

引用 0|浏览0
暂无评分
摘要
Memristors in crossbar arrays can accomplish computing operations while storing data at the same physical location, enabling a cost-efficient latency-free solution to the von Neumann bottleneck. Amorphous oxide semiconductor (AOS)-based memristors can be engineered to perform filamentary- and/or interface-type resistive switching. Their superior characteristics such as high flexibility compatible with low-temperature and easy manufacturing evidence their potential for embedded flexible neuromorphic technologies. In this chapter, the state-of-the-art on AOS-based resistive switching devices is analysed, along with their suitability for specific neuromorphic applications such as in-memory computation and deep and spiking neural networks. Currently, crosstalk is the main obstacle to large-scale crossbar integration and, therefore, the proposed main approaches to overcome this obstacle are discussed. Here, given the high level of behaviour control offered by AOS-based memristors, self-rectifying characteristics or optoelectronic features can be established. Moreover, the compatibility of AOS films with both memristors and thin-film transistors provides the necessary means for active crossbars to be developed in a cost-efficient, simple and higher-interconnectivity manner.
更多
查看译文
关键词
oxide,brain-inspired
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要