Design and Simulation of a Novel Multi-Floating-gate Synaptic Nanowire Transistor for Neuromorphic Computing

Xiaokang Li,Yuancheng Yang,Gong Chen,Shuang Sun,Qifeng Cai,Xiaoqiao Dong,Xiaoyan Xu,Xia An,Ming Li,Ru Huang
DOI: https://doi.org/10.1109/icsict.2018.8565016
2018-01-01
Abstract: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.
What problem does this paper attempt to address?