Integrated Artificial Neurons from Metal Halide Perovskites

TL;DR

This paper presents a novel on-chip artificial neuron using halide perovskite memristive devices, demonstrating energy-efficient stochastic leaky integrate-and-fire behavior suitable for neural network integration.

Contribution

It introduces a fully on-chip artificial neuron based on halide perovskite memristors, showcasing stochastic firing and low energy consumption, advancing memristive neural hardware.

Findings

Demonstrated stochastic leaky integrate-and-fire behavior with 20-60 pJ energy per spike.

Achieved lower energy consumption than biological neurons.

Enabled integration with halide perovskite synapses for energy-efficient neural networks.

Abstract

Hardware neural networks could perform certain computational tasks orders of magnitude more energy-efficiently than conventional computers. Artificial neurons are a key component of these networks and are currently implemented with electronic circuits based on capacitors and transistors. However, artificial neurons based on memristive devices are a promising alternative, owing to their potentially smaller size and inherent stochasticity. But despite their promise, demonstrations of memristive artificial neurons have so far been limited. Here we demonstrate a fully on-chip artificial neuron based on microscale electrodes and halide perovskite semiconductors as the active layer. By connecting a halide perovskite memristive device in series with a capacitor, the device demonstrates stochastic leaky integrate-and-fire behavior, with an energy consumption of 20 to 60 pJ per spike, lower than that of a biological neuron. We simulate populations of our neuron and show that the stochastic firing allows the detection of sub-threshold inputs. The neuron can easily be integrated with previously-demonstrated halide perovskite artificial synapses in energy-efficient neural networks.