Abstract:Inspired by the human brain, the spike-based neuromorphic system has attracted strong research enthusiasm because of the high energy efficiency and powerful computational capability, in which the spiking neurons and plastic synapses are two fundamental building blocks. Recently, two-terminal threshold switching (TS) devices have been regarded as promising candidates for building spiking neurons in hardware. However, how circuit parameters affect the spiking behavior of TS-based neurons is still an open question. Here, based on a leaky integrate-and-fire (LIF) neuron circuit, we systematically study the effect of both the extrinsic and intrinsic factors of NbO x -based TS neurons on their spiking behaviors. The extrinsic influence factors contain input intensities, connected synaptic weights, and parallel capacitances. To illustrate the effect of intrinsic factors, including the threshold voltage, holding voltage, and high/low resistance states of NbO x devices, we propose an empirical model of the fabricated NbO x devices, fitting well with the experimental results. The results indicate that with enhancing the input intensity, the spiking frequency increases first then decreases after reaching a peak value. Except for the connected synaptic weights, all other parameters can modulate the spiking peak frequency under high enough input intensity. Also, the relationship between energy consumption per spike and frequency of the neuron cell is further studied, leading guidance to design neuron circuits in a system to obtain the lowest energy consumption. At last, to demonstrate the practical applications of TS-based neurons, we construct a spiking neural network (SNN) to control the cart-pole using reinforcement learning, obtaining a reward score up to 450. This work provides valuable guidance on building compact LIF neurons based on TS devices and further bolsters the construction of high-efficiency neuromorphic systems.

A Scalable Artificial Neuron Based on Ultrathin Two-Dimensional Titanium Oxide

Flexible, Transparent And Wafer‐scale Artificial Synapse Array Based on TiOx/Ti3C2Tx Film for Neuromorphic Computing

Artificial Neurosynaptic Device Based on Amorphous Oxides for Artificial Neural Network Constructing

Compact artificial neuron based on anti-ferroelectric transistor

Application of artificial synapse based on Al-doped SrTiO 3 thin film in neuromorphic computing

An artificial synapse based on Sr(Ti, Co)O3 films

Low-voltage solution-processed artificial optoelectronic hybrid-integrated neuron based on 2D MXene for multi-task spiking neural network

Salt-assisted vapor–liquid–solid growth of high-quality ultrathin nickel oxide flakes for artificial synapses in image recognition applications

Spiking Neural Network Integrated with Impact Ionization Field-Effect Transistor Neuron and a Ferroelectric Field-Effect Transistor Synapse

Oxygen vacancies modulated VO2 for neurons and Spiking Neural Network construction

Artificial synapse arrays based on SiOx/TiOx memristive crossbar with high uniformity for neuromorphic computing

An Artificial Neural Network Based on Oxide Synaptic Transistor for Accurate and Robust Image Recognition

High-Yield and Uniform NbO x -Based Threshold Switching Devices for Neuron Applications

High-Yield and Uniform NbOx-Based Threshold Switching Devices for Neuron Applications

Engineering Spiking Neurons Using Threshold Switching Devices for High-Efficient Neuromorphic Computing

Spiking neurons with spatiotemporal dynamics and gain modulation for monolithically integrated memristive neural networks

Dielectric Engineered Two-Dimensional Neuromorphic Transistors

Memristive artificial synapses based on Au–TiO2 composite thin film for neuromorphic computing

Quasi-two-dimensional α-molybdenum oxide thin film prepared by magnetron sputtering for neuromorphic computing

Taming Prolonged Ionic Drift–Diffusion Dynamics for Brain‐Inspired Computation