Search

With the booming growth of artificial intelligence (AI), the traditional von Neumann computing architecture based on complementary metal oxide semiconductor devices are facing memory wall and power wall. Memristor based in-memory computing can potentially overcome the current bottleneck of computer and achieve hardware breakthrough. In this review, the recent progress of memory devices in material and structure design, device performance and applications are summarized. Various resistive switching materials, including electrodes, binary oxides, perovskites, organics, and two-dimensional materials, are presented and their role in the memristor are discussed. Subsequently, the construction of shaped electrodes, the design of functional layer and other factors influencing the device performance are analyzed. We focus on the modulation of the resistances and the effective methods to enhance the performance. Furthermore, synaptic plasticity, optical-electrical properties, the fashionable applications in logic operation and analog calculation are introduced. Finally, some critical issues such as the resistive switching mechanism, multi-sensory fusion, system-level optimization are discussed. [ABSTRACT FROM AUTHOR]

As a key building block of the biological cortex, synapses are powerful information processing units that enable highly complex nonlinear computations. The realization of artificial synapses with similar capabilities has important implications for building intelligent, neuromorphic systems. Here, we demonstrate an artificial synapse based on NbOx nonvolatile memristor to mimic multifunctional bionic applications such as nociceptor and associative learning. Combined experimental characterization with COMSOL simulation, the traditional resistance switching characteristics, which are the decisive factor for the synapse properties are in-depth analyzed. It can be proposed that the I–V characteristics of Pt/NbOx/TiN memristor are governed by core-shell filaments consisting of the shell region of sub-stoichiometric Nb2O5−δ and the core of NbO2. On the basis of the core-shell filament model, it can be reasonably explained that Ohmic conduction and Poole-Frenkel conduction take turns to dominate the current flowing in the memristive device, leading to the zigzag evolution of current during the operation process of NbOx-based device. The simulations of synaptic plasticity, including long-term potentiation/depression (LTP/LTD), paired-pulse facilitation (PPF), and spike-timing-dependent plasticity (STDP), exhibiting that the NbOx can be utilized for an artificial synapse. Furthermore, bionic functions such as hyperalgesia and allodynia of a nociceptor and a series of associative learning behaviors in Pavlovian dog experiment are mimicked, illustrating that the Pt/NbOx/TiN have great potential for highly simplified artificial neural network applications. [ABSTRACT FROM AUTHOR]

Flexible memristor is one of the most promising wearable devices for abundant data storage and processing. In this work, interface engineering by inserting the Al2O3 barrier layer is carried out to construct Pt/TiOx/Al2O3/Pt/indium tin oxide (ITO) flexible artificial synapse device. The memristor performance can be maintained even under 1000 times of bending without degradation, demonstrating its excellent mechanical property. With the Al2O3 diffusion barrier layer, the oxygen vacancies (${V}_{\text {o}}$) movement is slowed down for filaments formation and rupture, thus it boosts up the synaptic plasticity, including long-term potentiation/depression, paired-pulse facilitation (PPF), and spike-timing-dependent plasticity (STDP). Moreover, on the basis of the enhanced symmetry and linearity of conductance for Pt/TiOx/Al2O3/Pt/indium tin oxide (ITO) memristor, the neural network simulation for supervised learning presents an online learning pattern recognition, the accuracy can achieve to 91.15%. Overall, the Pt/TiOx/Al2O3/Pt/ITO memristor with excellent flexibility is a promising emulator for biological synapses, which could be beneficial to future flexible memristor-based neuromorphic computing. [ABSTRACT FROM AUTHOR]

Neuromorphic devices are among the most emerging electronic components to realize artificial neural systems and replace traditional complementary metal–oxide semiconductor devices in recent times. In this work, tri‐layer HfO2/BiFeO3(BFO)/HfO2 memristors are designed by inserting traditional ferroelectric BFO layers measuring ≈4 nm after thickness optimization. The novel designed memristor shows excellent resistive switching (RS) performance such as a storage window of 104 and multi‐level storage ability. Remarkably, essential synaptic functions can be successfully realized on the basis of the linearity of conductance modulation. The pattern recognition simulation based on neuromorphic network is conducted with 91.2% high recognition accuracy. To explore the RS performance enhancement and artificial synaptic behaviors, conductive filaments (CFs) composed of Hafnium (Hf) single crystal with a hexaganal lattice structure are observed using high‐resolution transmission electron microscopy. It is reasonable to believe that the sufficient oxygen vacancies in the inserting BFO thin film play a crucial role in adjusting the morphology and growth of Hf CFs, which leads to the promising synaptic and enhanced RS behavior, thus demonstrating the potential of this memristor for use in neuromorphic computing. [ABSTRACT FROM AUTHOR]