Low‐Voltage Oscillatory Neurons for Memristor‐Based Neuromorphic Systems

Qilin Hua, Huaqiang Wu, Bin Gao, Qingtian Zhang, Wei Wu, Yujia Li, Xiaohu Wang, Weiguo Hu, He Qian

Global Challenges · 2019 · 68 citations · 29 references

DOIFull text

Open access

Concepts

TL;DR

Neuromorphic systems aim to emulate biological neural networks, but conventional artificial neurons require high operating voltages and exhibit large leakage currents, leading to excessive power consumption. This work introduces a low‑voltage oscillatory neuron based on an Ag filamentary threshold‑switching memristor that operates below 0.6 V and consumes less than 1.8 µW. The neuron employs the memristor’s threshold switching to generate leaky integrate‑and‑fire dynamics and threshold‑driven spiking output. It achieves high endurance (>10 T) while reliably producing leaky integrate‑and‑fire and spiking behavior.

Abstract

Neuromorphic systems consisting of artificial neurons and synapses can process complex information with high efficiency to overcome the bottleneck of von Neumann architecture. Artificial neurons are essentially required to possess functions such as leaky integrate-and-fire and output spike. However, previous reported artificial neurons typically have high operation voltage and large leakage current, leading to significant power consumption, which is contrary to the energy-efficient biological model. Here, an oscillatory neuron based on Ag filamentary threshold switching memristor (TS) that has a low operation voltage (<0.6 V) with ultralow power consumption (<1.8 µW) is presented. It can trigger neuronal functions, including leaky integrate-and-fire and threshold-driven spiking output, with high endurance (>10

References

29