Neuroinspired unsupervised learning and pruning with subquantum CBRAM arrays

Reliable Low-Current and Multilevel Memristive Electrochemical Neuromorphic Devices with Semi-Metal Sb Filament

Memristors are used in artificial neural networks owing to their exceptional integration capabilities and scalability. However, traditional memristors are hampered by limited resistance states and randomness, which curtails their application. The migration of metal ions critically influences the number of conductance states and the linearity of weight updates. Semi-metal filaments can provide subquantum conductance changes to the memristors due to the smaller single-atom conductance, such as Sb (≈0.01 G0 = 7.69 × 10-7 S). Here, a memristor featuring an active electrode composed of semi-metal Sb is introduced for the first time. This memristor demonstrates precise conductance control, a large on/off ratio, consistent switching, and prolonged retention exceeding 105 s. Density functional theory (DFT) calculations and characterization methods reveal the formation of Sb filaments during a set process. The interaction between Sb and O within the dielectric layer facilitates the Sb filaments' ability to preserve their morphology in the absence of electric fields.

Multi-level, forming and filament free, bulk switching trilayer RRAM for neuromorphic computing at the edge

CMOS-RRAM integration holds great promise for low energy and high throughput neuromorphic computing. However, most RRAM technologies relying on filamentary switching suffer from variations and noise, leading to computational accuracy loss, increased energy consumption, and overhead by expensive program and verify schemes. We developed a filament-free, bulk switching RRAM technology to address these challenges. We systematically engineered a trilayer metal-oxide stack and investigated the switching characteristics of RRAM with varying thicknesses and oxygen vacancy distributions to achieve reliable bulk switching without any filament formation. We demonstrated bulk switching at megaohm regime with high current nonlinearity, up to 100 levels without compliance current. We developed a neuromorphic compute-in-memory platform and showcased edge computing by implementing a spiking neural network for an autonomous navigation/racing task. Our work addresses challenges posed by existing RRAM technologies and paves the way for neuromorphic computing at the edge under strict size, weight, and power constraints.

Highly Reliable 3D Channel Memory and Its Application in a Neuromorphic Sensory System for Hand Gesture Recognition

Brain-inspired neuromorphic computing systems, based on a crossbar array of two-terminal multilevel resistive random-access memory (RRAM), have attracted attention as promising technologies for processing large amounts of unstructured data. However, the low reliability and inferior conductance tunability of RRAM, caused by uncontrollable metal filament formation in the uneven switching medium, result in lower accuracy compared to the software neural network (SW-NN). In this work, we present a highly reliable CoOx-based multilevel RRAM with an optimized crystal size and density in the switching medium, providing a three-dimensional (3D) grain boundary (GB) network. This design enhances the reliability of the RRAM by improving the cycle-to-cycle endurance and device-to-device stability of the I-V characteristics with minimal variation. Furthermore, the designed 3D GB-channel RRAM (3D GB-RRAM) exhibits excellent conductance tunability, demonstrating high symmetricity (624), low nonlinearity (βLTP/βLTD ∼ 0.20/0.39), and a large dynamic range (Gmax/Gmin ∼ 31.1). The cyclic stability of long-term potentiation and depression also exceeds 100 cycles (105 voltage pulses), and the relative standard deviation of Gmax/Gmin is only 2.9%. Leveraging these superior reliability and performance attributes, we propose a neuromorphic sensory system for finger motion tracking and hand gesture recognition as a potential elemental technology for the metaverse. This system consists of a stretchable double-layered photoacoustic strain sensor and a crossbar array neural network. We perform training and recognition tasks on ultrasonic patterns associated with finger motion and hand gestures, attaining a recognition accuracy of 97.9% and 97.4%, comparable to that of SW-NN (99.8% and 98.7%).

Emerging memristive artificial neuron and synapse devices for the neuromorphic electronics era

Growth of data eases the way to access the world but requires increasing amounts of energy to store and process. Neuromorphic electronics has emerged in the last decade, inspired by biological neurons and synapses, with in-memory computing ability, extenuating the 'von Neumann bottleneck' between the memory and processor and offering a promising solution to reduce the efforts both in data storage and processing, thanks to their multi-bit non-volatility, biology-emulated characteristics, and silicon compatibility. This work reviews the recent advances in emerging memristive devices for artificial neuron and synapse applications, including memory and data-processing ability: the physics and characteristics are discussed first, i.e., valence changing, electrochemical metallization, phase changing, interfaced-controlling, charge-trapping, ferroelectric tunnelling, and spin-transfer torquing. Next, we propose a universal benchmark for the artificial synapse and neuron devices on spiking energy consumption, standby power consumption, and spike timing. Based on the benchmark, we address the challenges, suggest the guidelines for intra-device and inter-device design, and provide an outlook for the neuromorphic applications of resistive switching-based artificial neuron and synapse devices.

Compute in-Memory with Non-Volatile Elements for Neural Networks: A Review from a Co-Design Perspective

Deep learning has become ubiquitous, touching daily lives across the globe. Today, traditional computer architectures are stressed to their limits in efficiently executing the growing complexity of data and models. Compute-in-memory (CIM) can potentially play an important role in developing efficient hardware solutions that reduce data movement from compute-unit to memory, known as the von Neumann bottleneck. At its heart is a cross-bar architecture with nodal non-volatile-memory elements that performs an analog multiply-and-accumulate operation, enabling the matrix-vector-multiplications repeatedly used in all neural network workloads. The memory materials can significantly influence final system-level characteristics and chip performance, including speed, power, and classification accuracy. With an over-arching co-design viewpoint, this review assesses the use of cross-bar based CIM for neural networks, connecting the material properties and the associated design constraints and demands to application, architecture, and performance. Both digital and analog memory are considered, assessing the status for training and inference, and providing metrics for the collective set of properties non-volatile memory materials will need to demonstrate for a successful CIM technology.

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

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.

Energy Efficient Neuro-Inspired Phase-Change Memory Based on Ge4 Sb6 Te7 as a Novel Epitaxial Nanocomposite

Phase-change memory (PCM) is a promising candidate for neuro-inspired, data-intensive artificial intelligence applications, which relies on the physical attributes of PCM materials including gradual change of resistance states and multilevel operation with low resistance drift. However, achieving these attributes simultaneously remains a fundamental challenge for PCM materials such as Ge₂ Sb₂ Te₅ , the most commonly used material. Here bi-directional gradual resistance changes with ≈10× resistance window using low energy pulses are demonstrated in nanoscale PCM devices based on Ge₄ Sb₆ Te₇ , a new phase-change nanocomposite material . These devices show 13 resistance levels with low resistance drift for the first 8 levels, a resistance on/off ratio of ≈1000, and low variability. These attributes are enabled by the unique microstructural and electro-thermal properties of Ge₄ Sb₆ Te₇ , a nanocomposite consisting of epitaxial SbTe nanoclusters within the Ge-Sb-Te matrix, and a higher crystallization but lower melting temperature than Ge₂ Sb₂ Te₅ . These results advance the pathway toward energy-efficient analog computing using PCM.

Neuroscience and Network Dynamics Toward Brain-Inspired Intelligence

This article surveys the interdisciplinary research of neuroscience, network science, and dynamic systems, with emphasis on the emergence of brain-inspired intelligence. To replicate brain intelligence, a practical way is to reconstruct cortical networks with dynamic activities that nourish the brain functions, instead of using only artificial computing networks. The survey provides a complex network and spatiotemporal dynamics (abbr. network dynamics) perspective for understanding the brain and cortical networks and, furthermore, develops integrated approaches of neuroscience and network dynamics toward building brain-inspired intelligence with learning and resilience functions. Presented are fundamental concepts and principles of complex networks, neuroscience, and hybrid dynamic systems, as well as relevant studies about the brain and intelligence. Other promising research directions, such as brain science, data science, quantum information science, and machine behavior are also briefly discussed toward future applications.

A Learning-Rate Modulable and Reliable TiOx Memristor Array for Robust, Fast, and Accurate Neuromorphic Computing

Realization of memristor-based neuromorphic hardware system is important to achieve energy efficient bigdata processing and artificial intelligence in integrated device system-level. In this sense, uniform and reliable titanium oxide (TiO_x ) memristor array devices are fabricated to be utilized as constituent device element in hardware neural network, representing passive matrix array structure enabling vector-matrix multiplication process between multisignal and trained synaptic weight. In particular, in situ convolutional neural network hardware system is designed and implemented using a multiple 25 × 25 TiO_x memristor arrays and the memristor device parameters are developed to bring global constant voltage programming scheme for entire cells in crossbar array without any voltage tuning peripheral circuit such as transistor. Moreover, the learning rate modulation during in situ hardware training process is successfully achieved due to superior TiO_x memristor performance such as threshold uniformity (≈2.7%), device yield (> 99%), repetitive stability (≈3000 spikes), low asymmetry value of ≈1.43, ambient stability (6 months), and nonlinear pulse response. The learning rate modulable fast-converging in situ training based on direct memristor operation shows five times less training iterations and reduces training energy compared to the conventional hardware in situ training at ≈95.2% of classification accuracy.

Organismic Memristive Structures With Variable Functionality for Neuroelectronics

In this paper, we report an approach to design nanolayered memristive compositions based on TiO2/Al2O3 bilayer structures with analog non-volatile and volatile tuning of the resistance. The structure of the TiO2 layer drives the physical mechanism underlying the non-volatile resistance switching, which can be changed from electronic to ionic, enabling the synaptic behavior emulation. The presence of the anatase phase in the amorphous TiO2 layer induces the resistive switching mechanism due to electronic processes. In this case, the switching of the resistance within the range of seven orders of magnitude is experimentally observed. In the bilayer with amorphous titanium dioxide, the participation of ionic processes in the switching mechanism results in narrowing the tuning range down to 2–3 orders of magnitude and increasing the operating voltages. In this way, a combination of TiO2/Al2O3 bilayers with inert electrodes enables synaptic behavior emulation, while active electrodes induce the neuronal behavior caused by cation density variation in the active Al2O3 layer of the structure. We consider that the proposed approach could help to explore the memristive capabilities of nanolayered compositions in a more functional way, enabling implementation of artificial neural network algorithms at the material level and simplifying neuromorphic layouts, while maintaining all benefits of neuromorphic architectures.

Hafnium Oxide (HfO2 ) - A Multifunctional Oxide: A Review on the Prospect and Challenges of Hafnium Oxide in Resistive Switching and Ferroelectric Memories

Hafnium oxide (HfO₂ ) is one of the mature high-k dielectrics that has been standing strong in the memory arena over the last two decades. Its dielectric properties have been researched rigorously for the development of flash memory devices. In this review, the application of HfO₂ in two main emerging nonvolatile memory technologies is surveyed, namely resistive random access memory and ferroelectric memory. How the properties of HfO₂ equip the former to achieve superlative performance with high-speed reliable switching, excellent endurance, and retention is discussed. The parameters to control HfO₂ domains are further discussed, which can unleash the ferroelectric properties in memory applications. Finally, the prospect of HfO₂ materials in emerging applications, such as high-density memory and neuromorphic devices are examined, and the various challenges of HfO₂ -based resistive random access memory and ferroelectric memory devices are addressed with a future outlook.

MoS2 Synapses with Ultra-low Variability and Their Implementation in Boolean Logic

Brain-inspired computing enabled by memristors has gained prominence over the years due to the nanoscale footprint and reduced complexity for implementing synapses and neurons. The demonstration of complex neuromorphic circuits using conventional materials systems has been limited by high cycle-to-cycle and device-to-device variability. Two-dimensional (2D) materials have been used to realize transparent, flexible, ultra-thin memristive synapses for neuromorphic computing, but with limited knowledge on the statistical variation of devices. In this work, we demonstrate ultra-low-variability synapses using chemical vapor deposited 2D MoS₂ as the switching medium with Ti/Au electrodes. These devices, fabricated using a transfer-free process, exhibit ultra-low variability in SET voltage, RESET power distribution, and synaptic weight update characteristics. This ultra-low variability is enabled by the interface rendered by a Ti/Au top contact on Si-rich MoS₂ layers of mixed orientation, corroborated by transmission electron microscopy (TEM), electron energy loss spectroscopy (EELS), and X-ray photoelectron spectroscopy (XPS). TEM images further confirm the stability of the device stack even after subjecting the device to 100 SET-RESET cycles. Additionally, we implement logic gates by monolithic integration of MoS₂ synapses with MoS₂ leaky integrate-and-fire neurons to show the viability of these devices for non-von Neumann computing.

Exploring Low Power and Ultrafast Memristor on p-Type van der Waals SnS

Memristor devices that exhibit high integration density, fast speed, and low power consumption are candidates for neuromorphic devices. Here, we demonstrate a filament-based memristor using p-type SnS as the resistive switching material, exhibiting superlative metrics such as a switching voltage ∼0.2 V, a switching speed faster than 1.5 ns, high endurance switching cycles, and an ultralarge on/off ratio of 10⁸. The device exhibits a power consumption as low as ∼100 fJ per switch. Chip-level simulations of the memristor based on 32 × 32 high-density crossbar arrays with 50 nm feature size reveal on-chip learning accuracy of 87.76% (close to the ideal software accuracy 90%) for CIFAR-10 image classifications. The ultrafast and low energy switching of p-type SnS compared to n-type transition metal dichalcogenides is attributed to the presence of cation vacancies and van der Waals gap that lower the activation barrier for Ag ion migration.

A new opportunity for the emerging tellurium semiconductor: making resistive switching devices

The development of the resistive switching cross-point array as the next-generation platform for high-density storage, in-memory computing and neuromorphic computing heavily relies on the improvement of the two component devices, volatile selector and nonvolatile memory, which have distinct operating current requirements. The perennial current-volatility dilemma that has been widely faced in various device implementations remains a major bottleneck. Here, we show that the device based on electrochemically active, low-thermal conductivity and low-melting temperature semiconducting tellurium filament can solve this dilemma, being able to function as either selector or memory in respective desired current ranges. Furthermore, we demonstrate one-selector-one-resistor behavior in a tandem of two identical Te-based devices, indicating the potential of Te-based device as a universal array building block. These nonconventional phenomena can be understood from a combination of unique electrical-thermal properties in Te. Preliminary device optimization efforts also indicate large and unique design space for Te-based resistive switching devices.

Resistive switching devices have great promise for a wide variety of technological applications. Here, Yang et al demonstrate that electrochemically induced tellurium filament can give rise to resistive switching, and show that devices based on this can provide a number of advantages compared to metallic filament-based devices.

Energy-efficient Mott activation neuron for full-hardware implementation of neural networks

To circumvent the von Neumann bottleneck, substantial progress has been made towards in-memory computing with synaptic devices. However, compact nanodevices implementing non-linear activation functions are required for efficient full-hardware implementation of deep neural networks. Here, we present an energy-efficient and compact Mott activation neuron based on vanadium dioxide and its successful integration with a conductive bridge random access memory (CBRAM) crossbar array in hardware. The Mott activation neuron implements the rectified linear unit function in the analogue domain. The neuron devices consume substantially less energy and occupy two orders of magnitude smaller area than those of analogue complementary metal-oxide semiconductor implementations. The LeNet-5 network with Mott activation neurons achieves 98.38% accuracy on the MNIST dataset, close to the ideal software accuracy. We perform large-scale image edge detection using the Mott activation neurons integrated with a CBRAM crossbar array. Our findings provide a solution towards large-scale, highly parallel and energy-efficient in-memory computing systems for neural networks.

Memristors With Controllable Data Volatility by Loading Metal Ion-Added Ionic Liquids

We demonstrate a new memristive device (IL-Memristor), in which an ionic liquid (IL) serve as a material to control the volatility of the resistance. ILs are ultra-low vapor pressure liquids consisting of cations and anions at room temperature, and their introduction into solid-state processes can provide new avenues in electronic device fabrication. Because the device resistance change in IL-Memristor is governed by a Cu filament formation/rupture in IL, we considered that the Cu filament stability affects the data retention characteristics. Therefore, we controlled the data retention time by clarifying the corrosion mechanism and performing the IL material design based on the results. It was found out that the corrosion of Cu filaments in the IL was ruled by the comproportionation reaction, and that the data retention characteristics of the devices varied depending on the valence of Cu ions added to the IL. Actually, IL-Memristors involving Cu(II) and Cu(I) show volatile and non-volatile nature with respect to the programmed resistance value, respectively. Our results showed that data volatility can be controlled through the metal ion species added to the IL. The present work indicates that IL-memristor is suitable for unique applications such as artificial neuron with tunable fading characteristics that is applicable to phenomena with a wide range of timescale.

Run-off election-based decision method for the training and inference process in an artificial neural network

Generally, the decision rule for classifying unstructured data in an artificial neural network system depends on the sequence results of an activation function determined by vector-matrix multiplication between the input bias signal and the analog synaptic weight quantity of each node in a matrix array. Although a sequence-based decision rule can efficiently extract a common feature in a large data set in a short time, it can occasionally fail to classify similar species because it does not intrinsically consider other quantitative configurations of the activation function that affect the synaptic weight update. In this work, we implemented a simple run-off election-based decision rule via an additional filter evaluation to mitigate the confusion from proximity of output activation functions, enabling the improved training and inference performance of artificial neural network system. Using the filter evaluation selected via the difference among common features of classified images, the recognition accuracy achieved for three types of shoe image data sets reached ~ 82.03%, outperforming the maximum accuracy of ~ 79.23% obtained via the sequence-based decision rule in a fully connected single layer network. This training algorithm with an independent filter can precisely supply the output class in the decision step of the fully connected network.

Neuromorphic Engineering: From Biological to Spike-Based Hardware Nervous Systems

The human brain is a sophisticated, high-performance biocomputer that processes multiple complex tasks in parallel with high efficiency and remarkably low power consumption. Scientists have long been pursuing an artificial intelligence (AI) that can rival the human brain. Spiking neural networks based on neuromorphic computing platforms simulate the architecture and information processing of the intelligent brain, providing new insights for building AIs. The rapid development of materials engineering, device physics, chip integration, and neuroscience has led to exciting progress in neuromorphic computing with the goal of overcoming the von Neumann bottleneck. Herein, fundamental knowledge related to the structures and working principles of neurons and synapses of the biological nervous system is reviewed. An overview is then provided on the development of neuromorphic hardware systems, from artificial synapses and neurons to spike-based neuromorphic computing platforms. It is hoped that this review will shed new light on the evolution of brain-like computing.

Emerging Memristive Artificial Synapses and Neurons for Energy-Efficient Neuromorphic Computing

Memristors have recently attracted significant interest due to their applicability as promising building blocks of neuromorphic computing and electronic systems. The dynamic reconfiguration of memristors, which is based on the history of applied electrical stimuli, can mimic both essential analog synaptic and neuronal functionalities. These can be utilized as the node and terminal devices in an artificial neural network. Consequently, the ability to understand, control, and utilize fundamental switching principles and various types of device architectures of the memristor is necessary for achieving memristor-based neuromorphic hardware systems. Herein, a wide range of memristors and memristive-related devices for artificial synapses and neurons is highlighted. The device structures, switching principles, and the applications of essential synaptic and neuronal functionalities are sequentially presented. Moreover, recent advances in memristive artificial neural networks and their hardware implementations are introduced along with an overview of the various learning algorithms. Finally, the main challenges of the memristive synapses and neurons toward high-performance and energy-efficient neuromorphic computing are briefly discussed. This progress report aims to be an insightful guide for the research on memristors and neuromorphic-based computing.

Vertical organic synapse expandable to 3D crossbar array

Recently, three-terminal synaptic devices have attracted considerable attention owing to their nondestructive weight-update behavior, which is attributed to the completely separated terminals for reading and writing. However, the structural limitations of these devices, such as a low array density and complex line design, are predicted to result in low processing speeds and high energy consumption of the entire system. Here, we propose a vertical three-terminal synapse featuring a remote weight update via ion gel, which is also extendable to a crossbar array structure. This synaptic device exhibits excellent synaptic characteristics, which are achieved via precise control of ion penetration onto the vertical channel through the weight-control terminal. Especially, the applicability of the developed vertical organic synapse array to neuromorphic computing is demonstrated using a simple crossbar synapse array. The proposed synaptic device technology is expected to be an important steppingstone to the development of high-performance and high-density neural networks.

Implementation of Dropout Neuronal Units Based on Stochastic Memristive Devices in Neural Networks with High Classification Accuracy

Neural networks based on memristive devices have achieved great progress recently. However, memristive synapses with nonlinearity and asymmetry seriously limit the classification accuracy. Moreover, insufficient number of training samples in many cases also have negative effect on the classification accuracy of neural networks due to overfitting. In this work, dropout neuronal units are developed based on stochastic volatile memristive devices of Ag/Ta2O5:Ag/Pt. The memristive neural network using the dropout neuronal units effectively solves the problem of overfitting and mitigates the negative effects of the nonideality of memristive synapses, eventually achieves a classification accuracy comparable to the theoretical limit. The stochastic and volatile switching performances of the Ag/Ta2O5:Ag/Pt device are attributed to the stochastical rupture of the Ag filament under high electrical stress in the Ta2O5 layer, according to the TEM observation and the kinetic Monte Carlo simulation.

Conductive-bridging random-access memories for emerging neuromorphic computing

With the increasing utilisation of artificial intelligence, there is a renewed demand for the development of novel neuromorphic computing owing to the drawbacks of the existing computing paradigm based on the von Neumann architecture. Extensive studies have been performed on memristors as their electrical nature is similar to those of biological synapses and neurons. However, most hardware-based artificial neural networks (ANNs) have been developed with oxide-based memristors owing to their high compatibility with mature complementary metal-oxide-semiconductor (CMOS) processes. Considering the advantages of conductive-bridging random-access memories (CBRAMs), such as their high scalability, high on-off current with a wide dynamic range, and low off-current, over oxide-based memristors, extensive studies on CBRAMs are required. In this review, the basics of operation of CBRAMs are examined in detail, from the formation of metal nanoclusters to filament bridging. Additionally, state-of-the-art experimental demonstrations of CBRAM-based artificial synapses and neurons are presented. Finally, CBRAM-based ANNs are discussed, including deep neural networks and spiking neural networks, along with other emerging computing applications. This review is expected to pave the way toward further development of large-scale CBRAM array systems.

Alloying conducting channels for reliable neuromorphic computing

A memristor¹ has been proposed as an artificial synapse for emerging neuromorphic computing applications^2,3. To train a neural network in memristor arrays, changes in weight values in the form of device conductance should be distinct and uniform³. An electrochemical metallization (ECM) memory^4,5, typically based on silicon (Si), has demonstrated a good analogue switching capability^6,7 owing to the high mobility of metal ions in the Si switching medium⁸. However, the large stochasticity of the ion movement results in switching variability. Here we demonstrate a Si memristor with alloyed conduction channels that shows a stable and controllable device operation, which enables the large-scale implementation of crossbar arrays. The conduction channel is formed by conventional silver (Ag) as a primary mobile metal alloyed with silicidable copper (Cu) that stabilizes switching. In an optimal alloying ratio, Cu effectively regulates the Ag movement, which contributes to a substantial improvement in the spatial/temporal switching uniformity, a stable data retention over a large conductance range and a substantially enhanced programmed symmetry in analogue conductance states. This alloyed memristor allows the fabrication of large-scale crossbar arrays that feature a high device yield and accurate analogue programming capability. Thus, our discovery of an alloyed memristor is a key step paving the way beyond von Neumann computing.

Research progress on solutions to the sneak path issue in memristor crossbar arrays

Since the emergence of memristors (or memristive devices), how to integrate them into arrays has been widely investigated. After years of research, memristor crossbar arrays have been proposed and realized with potential applications in nonvolatile memory, logic and neuromorphic computing systems. Despite the promising prospects of memristor crossbar arrays, one of the main obstacles for their development is the so-called sneak-path current causing cross-talk interference between adjacent memory cells and thus may result in misinterpretation which greatly influences the operation of memristor crossbar arrays. Solving the sneak-path current issue, the power consumption of the array will immensely decrease, and the reliability and stability will simultaneously increase. In order to suppress the sneak-path current, various solutions have been provided. So far, some reviews have considered some of these solutions and established a sophisticated classification, including 1D1M, 1T1M, 1S1M (D: diode, M: memristor, T: transistor, S: selector), self-selective and self-rectifying memristors. Recently, a mass of studies have been additionally reported. This review thus attempts to provide a survey on these new findings, by highlighting the latest research progress realized for relieving the sneak-path issue. Here, we first present the concept of the sneak-path current issue and solutions proposed to solve it. Consequently, we select some typical and promising devices, and present their structures and properties in detail. Then, the latest research activities focusing on single-device structures are introduced taking into account the mechanisms underlying these devices. Finally, we summarize the properties and perspectives of these solutions.

Controlling Cu Migration on Resistive Switching, Artificial Synapse, and Glucose/Saliva Detection by Using an Optimized AlO x Interfacial Layer in a-CO x -Based Conductive Bridge Random Access Memory

The Cu migration is controlled by using an optimized AlO _x interfacial layer, and effects on resistive switching performance, artificial synapse, and human saliva detection in an amorphous-oxygenated-carbon (a-CO _x )-based CBRAM platform have been investigated for the first time. The 4 nm-thick AlO _x layer in the Cu/AlO _x /a-CO _x /TiN _x O _y /TiN structure shows consecutive >2000 DC switching, tight distribution of SET/RESET voltages, a long program/erase (P/E) endurance of >10⁹ cycles at a low operation current of 300 μA, and artificial synaptic characteristics under a small pulse width of 100 ns. After a P/E endurance of >10⁸ cycles, the Cu migration is observed by both ex situ high-resolution transmission electron microscopy and energy-dispersive X-ray spectroscopy mapping images. Furthermore, the optimized Cu/AlO _x /a-CO _x /TiN _x O _y /TiN CBRAM detects glucose with a low concentration of 1 pM, and real-time measurement of human saliva with a small sample volume of 1 μL is also detected repeatedly in vitro. This is owing to oxidation-reduction of Cu electrode, and the switching mechanism is explored. Therefore, this CBRAM device is beneficial for future artificial intelligence application.

Gate-Tunable Synaptic Dynamics of Ferroelectric-Coupled Carbon-Nanotube Transistors

Artificial neural networks (ANNs) based on synaptic devices, which can simultaneously perform processing and storage of data, have superior computing performance compared to conventional von Neumann architectures. Here, we present a ferroelectric coupled artificial synaptic device with reliable weight update and storage properties for ANNs. The artificial synaptic device, which is based on a ferroelectric polymer capacitively coupled with an oxide dielectric via an electric-field-permeable, semiconducting single-walled carbon-nanotube channel, is successfully fabricated by inkjet printing. By controlling the ferroelectric polarization, synaptic dynamics, such as excitatory and inhibitory postsynaptic currents and long-term potentiation/depression characteristics, is successfully implemented in the artificial synaptic device. Furthermore, the constructed ANN, which is designed in consideration of the device-to-device variation within the synaptic array, efficiently executes the tasks of learning and recognition of the Modified National Institute of Standards and Technology numerical patterns.

Tin oxide artificial synapses for low power temporal information processing

Lateral memristors configured with inert Pt contacts and mixed phase tin oxide layers have exhibited immediate, forming-free, low-power bidirectional resistance switching. Activity dependent conductance and relaxation in the low resistance state resembled short term potentiation in biological synapses. After scanning probe microscopy, x-ray photoelectron spectroscopy and electrical measurements, the device characteristics were attributed to Joule heating induced decomposition of the minority SnO phase and formation of a SnO₂ conducting filament with higher effective n-type doping. Finally, the devices recognized input voltage pulse sequences and spectral data by returning unique conductance states, suggesting suitability for bio-inspired pattern recognition systems.

A Soft-Pruning Method Applied During Training of Spiking Neural Networks for In-memory Computing Applications

Inspired from the computational efficiency of the biological brain, spiking neural networks (SNNs) emulate biological neural networks, neural codes, dynamics, and circuitry. SNNs show great potential for the implementation of unsupervised learning using in-memory computing. Here, we report an algorithmic optimization that improves energy efficiency of online learning with SNNs on emerging non-volatile memory (eNVM) devices. We develop a pruning method for SNNs by exploiting the output firing characteristics of neurons. Our pruning method can be applied during network training, which is different from previous approaches in the literature that employ pruning on already-trained networks. This approach prevents unnecessary updates of network parameters during training. This algorithmic optimization can complement the energy efficiency of eNVM technology, which offers a unique in-memory computing platform for the parallelization of neural network operations. Our SNN maintains ~90% classification accuracy on the MNIST dataset with up to ~75% pruning, significantly reducing the number of weight updates. The SNN and pruning scheme developed in this work can pave the way toward applications of eNVM based neuro-inspired systems for energy efficient online learning in low power applications.