A Self-Rectifying Synaptic Memristor Array with Ultrahigh Weight Potentiation Linearity for a Self-Organizing-Map Neural Network

Abstract

Two-terminal self-rectifying (SR)-synaptic memristors are preeminent candidates for high-density and efficient neuromorphic computing, especially for future three-dimensional integrated systems, which can self-suppress the sneak path current in crossbar arrays. However, SR-synaptic memristors face the critical challenges of nonlinear weight potentiation and steep depression, hindering their application in conventional artificial neural networks (ANNs). Here, a SR-synaptic memristor (Pt/NiO x / WO 3−x :Ti/W) and cross-point array with sneak path current suppression features and ultrahigh-weight potentiation linearity up to 0.9997 are introduced. The image contrast enhancement and background filtering are demonstrated on the basis of the device array. Moreover, an unsupervised self-organizing map (SOM) neural network is first developed for orientation recognition with high recognition accuracy (0.98) and training efficiency and high resilience toward both noises and steep synaptic depression. These results solve the challenges of SR memristors in the conventional ANN, extending the possibilities of large-scale oxide SR-synaptic arrays for high-density, efficient, and accurate neuromorphic computing.

Porous crystalline materials for memories and neuromorphic computing systems

“…According to the near-sensor and in-sensor computing strategy, a low-level sensory processing system with fast response, small size, high efficiency, and low energy consumption can be achieved by constructing a “sensor-synapse” structure or developing an artificial synapse with sensing properties at the device level. 494,556,557,561–564 On the contrary, the high-level sensory processing always relies on array-level and ANN techniques, such as CNNs, deep neural networks (DNNs), and SNNs, to carry out cognitive tasks like classification, recognition, and localization. 494,560,562,563,565,566 Unfortunately, the high-level sensory processing in near-sensor and in-sensor computing systems still relies on the von Neumann computing architecture up to this point.…”

Implementation of Convolutional Neural Networks in Memristor Crossbar Arrays with Binary Activation and Weight Quantization

“…However, reliability issues persist in memristive devices, particularly concerning device-to-device (D2D) or cycle-to-cycle (C2C) variations arising due to the switching mechanism of memristive devices, which is predominantly based on the soft breakdown of dielectric layers. − Such variations can significantly degrade the performance of neural networks as inaccurate device states hinder precise VMM operations and lead to computation errors. Although a memristor array structure with active devices such as a transistor has been demonstrated to improve cell selectivity, − a passive crossbar array without active devices is advantageous for high-density integration of 4F 2 , thanks to the cross-point array structure. − With increasing cell resistance to mitigate IR drop caused by line resistances, the sneak path current issue in the passive crossbar array can be suppressed by designing bias schemes such as half-V and third-V schemes. , While the sneak path issue can be effectively suppressed by self-rectifying or monolithic integrable selectors, − it becomes negligible in a passive crossbar array as well during VMM operations because all the WLs and bitlines (BLs) are connected to known potential values, unlike stand-alone memory operations where most devices are on unselected WLs and BLs …”

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

“…Under intended stimulation, synaptic plasticity-which is classified into short-term plasticity (STP) and long-term plasticity (LTP)-allows for effective adjustment of the strength of connections between neurons. [43] Previous research has investigated and developed artificial synaptic devices, such as metal-oxide semiconductors, chalcogenides, and nanoscale two-terminal and three-terminal synaptic memristors with STP and LTP functionalities. [44] The biological mechanism of the artificial synaptic devices is assumed to be the controlled movement of ions in response to electrical signals, which results in nonlinear conductance changes in post-synaptic currents.…”