Abstract:Memristors which have the potential to continuing Moore's law gradually attract attention for their advantages in storage density and computing speed. In this work, a HfO2/WO3 heterojunction‐based memristor with multilevel resistive switching and excellent neuromorphic performances is proposed. In addition, the memristor‐based multilayer perception neural network, which shows high recognition accuracy to handwritten digits, is constructed and verified. With the boom of artificial intelligence (AI) and big data, electronics demand faster computing speed and lower power consumption, however, von Neumann architecture of current devices feature severe drawbacks for the further improvement of computing capability due to its design with separated memory and central processing unit (CPU). Fortunately, emerging nonvolatile memory devices, especially memristors, exhibit tremendous advantages in breaking the "memory wall" between memory and CPU by virtue of their in‐computing and neuromorphic computing abilities. Here, a WO3/HfO2 heterojunction‐based memristor is proposed, and the device exhibits extraordinary resistive switching behaviors (e.g., high ON/OFF ratio, stable endurance, long retention time, and multilevel resistance states) and neuromorphic characteristics (long‐term/short‐term synaptic activities). Further, the mechanism underlying the electrical performances of this device is studied. Silver conductive filaments and Schottky barrier models are proposed and explained successfully. Additionally, a multilayer layer perceptron neural network is constructed in terms of the memristor model, and variables embracing learning rate, algorithm, and training epochs, are explored to enhance the recognition accuracy of the network. Undoubtedly, the proposed high‐quality WO3/HfO2 heterojunction‐based memristor contributes to promoting the development of high‐density storage and neuromorphic computing technology, showing fascinating prospects in the era of AI.

BEOL Large-Scale Integration and Precise Programming of HfO$_{\textit{x}}$/AlO$_{\textit{y}}$ Superlattice-Like Multilevel Memristors

Self-rectifying Memristors with High Rectification Ratio and Dynamic Linearity for In-Memory Computing

HfOx/AlOy Superlattice‐Like Memristive Synapse

Stepped Identical Reset-Pulse and Lookup Table Programming of Multilevel Conductance in HfOx/AlOy Superlattice-Like Memristors

Performance Optimization of Atomic Layer Deposited HfO x Memristor by Annealing With Back-End-of-Line Compatibility

Performance Optimization of Atomic Layer Deposited HfOx Memristor by Annealing with Back-End-of-Line Compatibility

A Multi-level Memristor Based on Al-Doped HfO2 Thin Film.

Designing High‐Performance Storage in HfO2/BiFeO3 Memristor for Artificial Synapse Applications

Research on Electronic Synaptic Simulation of HfO2-based Memristor by Embedding Al2O3

Homo-layer hafnia-based memristor with large analog switching window

HfO2/WO3 Heterojunction Structured Memristor for High‐Density Storage and Neuromorphic Computing

Thousands of conductance levels in memristors monolithically integrated on CMOS

Diverse long-term potentiation and depression based on multilevel LiSiOx memristor for neuromorphic computing

BiFeO3/SrTiO3 Superlattice-like Based Ferroelectric Memristors with Pronounced Artificial Synaptic Plasticity

Customized Binary and Multi-Level HfO2−x-based Memristors Tuned by Oxidation Conditions

Room-Temperature Fabricated Multilevel Nonvolatile Lead-Free Cesium Halide Memristors for Reconfigurable In-Memory Computing

One‐Dimensional Covalent Organic Framework‐Based Multilevel Memristors for Neuromorphic Computing

Diverse Long-Term Potentiation and Depression Based on Multilevel LiSiOxmemristor for Neuromorphic Computing.

HfOx/AlOy Superlattice‐Like Memristive Synapse (adv. Sci. 21/2022)

Passive LiNbO₃ Memristor With Multilevel States for Neuromorphic Computing

Hf0.5Zr0.5O2-based Ferroelectric Memristor with Multilevel Storage Potential and Artificial Synaptic Plasticity