Work

Threshold Switching Memristor Modeling for Spiking Neuron Design

Year: 2024

Type: article

Abstract: Spiking neurons, an essential building block in neuromorphic computing for encoding the input signals into spikes, require device advances and circuit designs for the realization. Recent studies show that the threshold switching memristors (TSMs) can enable spiking neurons with simple circuits and efficient data processing. Herein, we study and model the switching behavior of TSMs using the standardized Verilog-A language. The model facilitatesLeaky Integrate-and-Fire(LIF) neuron design on Cadence Virtuoso, notably, without the need for peripheral threshold spiking or resetting circuits. The neuron successfully performs input signal integration in both digital and analog forms for spike generation. Given the versability of the model, the study is expected to advance spiking neuron designs towards VLSI neuron circuits for neuromorphic computing.

Source: IEEE Electron Device Letters

DOI: 10.1109/led.2024.3424586

Authors: Pengyu Liu, Lekai Song, Kong-Pang Pun, Guohua Hu

Institution: Chinese University of Hong Kong

Language: English

Topic: Advanced Memory and Neural Computing

Subfield: Electrical and Electronic Engineering

Field: Engineering

Domain: Physical Sciences

Open Access status: hybrid

Estimated APC paid: $2,495

Funders: National Natural Science Foundation of China, Chinese University of Hong Kong