当前位置: X-MOL 学术Science › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
2D materials–based homogeneous transistor-memory architecture for neuromorphic hardware
Science ( IF 56.9 ) Pub Date : 2021-09-24 , DOI: 10.1126/science.abg3161
Lei Tong 1 , Zhuiri Peng 1 , Runfeng Lin 1 , Zheng Li 1 , Yilun Wang 1 , Xinyu Huang 1 , Kan-Hao Xue 1 , Hangyu Xu 2 , Feng Liu 3 , Hui Xia 2 , Peng Wang 2 , Mingsheng Xu 4 , Wei Xiong 1 , Weida Hu 2 , Jianbin Xu 5 , Xinliang Zhang 1 , Lei Ye 1 , Xiangshui Miao 1
Affiliation  

In neuromorphic hardware, peripheral circuits and memories based on heterogeneous devices are generally physically separated. Thus exploring homogeneous devices for these components is an important issue for improving module integration and resistance matching. Inspired by ferroelectric proximity effect on two-dimensional materials, we present a tungsten diselenide-on-LiNbO3 cascaded architecture as a basic device that functions as a nonlinear transistor, assisting the design of operational amplifiers for analog signal processing (ASP). It also functions as a nonvolatile memory cell, achieving memory operating (MO). Based on this homogeneous architecture, an analog signal processing-memory operating integrated system for binary classification and the design of ternary content-addressable memory were investigated for potential use in neuromorphic hardware.

中文翻译:

用于神经形态硬件的基于二维材料的同质晶体管内存架构

在神经形态硬件中,基于异构设备的外围电路和存储器通常是物理分离的。因此,为这些组件探索同质器件是提高模块集成度和电阻匹配的重要问题。受二维材料上的铁电邻近效应的启发,我们提出了一种二硒化钨-LiNbO 3级联架构作为一种基本器件,起到非线性晶体管的作用,有助于设计用于模拟信号处理 (ASP) 的运算放大器。它还用作非易失性存储单元,实现内存操作 (MO)。基于这种同构架构,研究了一种用于二进制分类的模拟信号处理-内存操作系统集成系统和三元内容可寻址内存的设计,以用于神经形态硬件。
更新日期:2021-09-10
down
wechat
bug