Definition of a Localized Conducting Path via Suppressed Charge Injection in Oxide Memristors for Stable Practical Hardware Neural Networks

Abstract

Oxide-based memristors have been demonstrated as suitable options for memory components in neuromorphic systems. In such devices, the resistive switching characteristics are caused by the formation of conductive filaments (CFs) comprising oxygen vacancies. Thus, the electrical performance is primarily governed by the CF structure. Despite various approaches for regulating the oxygen vacancy distributions in oxide memristors, controlling the CF structure without modifying the device configuration related to material compatibility is still a challenge. This study demonstrates an effective strategy for localizing CF distributions in memristors by suppressing charge injection during the formation of conducting paths. As the injected charge quantity is reduced in the electroforming process of the oxide memristor, the CF distributions become narrower, leading to more reproducible and stable resistive switching characteristics in the device. Based on these findings, a reliable hardware neural network comprising oxide memristors is constructed to recognize complex images. The developed memristor has been employed as a synaptic memory component in systems without degradation for a long time. This promising concept of oxide memristors acting as stable synaptic components holds great potential for developing practical neuromorphic systems and their expansion into artificial intelligent systems.

Filamentary-based organic memristors for wearable neuromorphic computing systems

“…In typical memristors, a sandwich structure comprising metal/switching layer/metal is employed, and the resistance state of the switching layer is adjusted by external electric stimuli. To date, various switching layers, including oxide films [20][21][22], 2D materials [23,24], and organic films [25][26][27], have been utilized in memristors. Among these, organic material-based memristors have been demonstrated as viable synaptic components for smart wearable computing systems, owing to their tunable electrical characteristics, mechanical flexibility, and biocompatibility [19,28,29].…”

Synaptic Characteristics and Vector‐Matrix Multiplication Operation in Highly Uniform and Cost‐Effective Four‐Layer Vertical RRAM Array

“…[17,18] It is time to begin considering a possible path for the commercialization of RRAM, which has attracted much interest in research laboratories and the semiconductor industry over the past years. [19][20][21][22][23][24][25][26][27] Resistive switching memories of various materials are applicable in commercialization in two ways: 1) high-density memory with the likely architecture of a crossbar array or vertical NAND-like structure [17] ; and 2) new computing or memory unit, such as in neuromorphic computing or dynamic logic software. [18] These ideas may coincide with the "beyond-Moore" approach.…”

“…There are two ideal methods for designing high-density 3D structures: lateral and vertical scaling. [24,25] Lateral scaling, also known as "3D X-point RRAM," employs stacked X-point arrays with metal lines orthogonally crossing one another, and memory cells are produced where the lines cross on the plane. [5,6,14] The memory scaling barrier may also be solved through vertical scaling technologies alongside lateral scaling.…”

“…Moreover, the simplified lithography process of VRRAM leads to more cost-effective manufacturing. [25] Resistive switching is observed in various materials, but transition metal oxides (TMO) have drawn the most attention due to their ease of fabrication and stable structure. [20] HfO x -based RRAM has differentiated itself from other alternatives by offering advantages such as low operation voltage, fast switching, and excellent thermal stability.…”