神经形态工程学
长时程增强
晶体管
计算机科学
CMOS芯片
突触重量
纳米线
电子工程
互连
人工神经网络
电气工程
材料科学
光电子学
工程类
计算机网络
人工智能
化学
生物化学
受体
电压
作者
Xiaokang Li,Yuancheng Yang,Gong Chen,Shuang Sun,Qifeng Cai,Xiaoqiao Dong,Xiaoyan Xu,Xia An,Ming Li,Ru Huang
标识
DOI:10.1109/icsict.2018.8565016
摘要
In this paper, we proposed a novel multi-floating-gate synaptic nanowire transistor which can emulate synapses behaviors such as long-term potentiation (LTP), long-term depression (LTD), integration signals from multiple pre- synapses and successive synaptic weight modulation, it can also realize high density interconnection in current CMOS IC technology to build complex neural network. Meanwhile, a novel scheme was also proposed to simulate amplitude-adjustable function by properly controlling the polarity-gates. Therefore, the multiple synaptic behaviors can be realized by the proposed new device, making it a competitive candidate for neuromorphic systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI