Bio-Inspired Neurons Based on Novel Leaky-FeFET with Ultra-Low Hardware Cost and Advanced Functionality for All-Ferroelectric Neural Network

Abstract

For the brain-inspired neuromorphic computing, various emerging memory devices, including FeFET, have been applied to develop the artificial synapses, while the artificial neurons are still mostly CMOS-implemented and suffer from high-hardware-cost issue, especially when expanding advanced functions. In this work, a novel leaky-FeFET (L-FeFET) based on partially crystallized Hf <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0.5</sub> Zr <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0.5</sub> O <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sub> layer is designed to mimic biological neurons. For the first time, we propose and experimentally demonstrate a capacitor-less L-FeFET neuron for basic leaky-integrate-and-fire function with ultra-low hardware cost of only one transistor and one resistor. Furthermore, a new hybrid L-FeFET-CMOS neuron is implemented to expand advanced spike-frequency adaption with almost half of hardware cost compared with CMOS neuron. This work provides a highly-integrated and inherently-low-energy implementation for neuron and the possibility for all-ferroelectric neural networks.