Training and Operation of an Integrated Neuromorphic Network Based on Metal-Oxide Memristors

Abstract

Despite all the progress of semiconductor integrated circuit technology, the extreme complexity of the human cerebral cortex makes the hardware implementation of neuromorphic networks with a comparable number of devices exceptionally challenging. One of the most prospective candidates to provide comparable complexity, while operating much faster and with manageable power dissipation, are so-called CrossNets based on hybrid CMOS/memristor circuits. In these circuits, the usual CMOS stack is augmented with one or several crossbar layers, with adjustable two-terminal memristors at each crosspoint. Recently, there was a significant progress in improvement of technology of fabrication of such memristive crossbars and their integration with CMOS circuits, including first demonstrations of their vertical integration. Separately, there have been several demonstrations of discrete memristors as artificial synapses for neuromorphic networks. Very recently such experiments were extended to crossbar arrays of phase-change memristive devices. The adjustment of such devices, however, requires an additional transistor at each crosspoint, and hence the prospects of their scaling are less impressive than those of metal-oxide memristors, whose nonlinear I-V curves enable transistor-free operation. Here we report the first experimental implementation of a transistor-free metal-oxide memristor crossbar with device variability lowered sufficiently to demonstrate a successful operation of a simple integrated neural network, a single layer-perceptron. The network could be taught in situ using a coarse-grain variety of the delta-rule algorithm to perform the perfect classification of 3x3-pixel black/white images into 3 classes. We believe that this demonstration is an important step towards the implementation of much larger and more complex memristive neuromorphic networks.

21 pages, 12 figures

References in corpus (1)

• Brain-like associative learning using a nanoscale non-volatile phase change synaptic device array

Cited by in corpus (36)

• A Survey of Neuromorphic Computing and Neural Networks in Hardware
• Magnetic skyrmion-based synaptic devices
• Mitigating Asymmetric Nonlinear Weight Update Effects in Hardware Neural Network based on Analog Resistive Synapse
• Stochastic Spiking Neural Networks Enabled by Magnetic Tunnel Junctions: From Nontelegraphic to Telegraphic Switching Regimes
• Delay dynamics of neuromorphic optoelectronic nanoscale resonators: Perspectives and applications
• A differential memristive synapse circuit for on-line learning in neuromorphic computing systems
• Synaptic dynamics in complex self-assembled nanoparticle networks
• Training a Probabilistic Graphical Model with Resistive Switching Electronic Synapses
• PUMA: A Programmable Ultra-efficient Memristor-based Accelerator for Machine Learning Inference
• Harnessing Intrinsic Noise in Memristor Hopfield Neural Networks for Combinatorial Optimization
• X-CHANGR: Changing Memristive Crossbar Mapping for Mitigating Line-Resistance Induced Accuracy Degradation in Deep Neural Networks
• Implementation of Multilayer Perceptron Network with Highly Uniform Passive Memristive Crossbar Circuits
• 4K-Memristor Analog-Grade Passive Crossbar Circuit
• Non-volatile Hierarchical Temporal Memory: Hardware for Spatial Pooling
• RESPARC: A Reconfigurable and Energy-Efficient Architecture with Memristive Crossbars for Deep Spiking Neural Networks
• 5 Parallel Prism: A topology for pipelined implementations of convolutional neural networks using computational memory
• Advancing Memristive Analog Neuromorphic Networks: Increasing Complexity, and Coping with Imperfect Hardware Components
• Gate-tunable, normally-on to normally-off memristance transition in patterned LaAlO3/SrTiO3 interfaces
• Transient dynamics of pulse-driven memristors in the presence of a stable fixed point
• A system of different layers of abstraction for artificial intelligence
• Oxide layer thickness effects on the resistance switching characteristics of Ti/TiO2-NT/Au structure
• Bridging the Gap Between Neural Networks and Neuromorphic Hardware with A Neural Network Compiler
• Charge-based computing with analogue reconfigurable gates
• Highly-Secure Physically Unclonable Cryptographic Primitives Using Nonlinear Conductance and Analog State Tuning in Memristive Crossbar Arrays
• Hardware-friendly Neural Network Architecture for Neuromorphic Computing
• MaD: Mapping and debugging framework for implementing deep neural network onto a neuromorphic chip with crossbar array of synapses
• A Physical Unclonable Function with Redox-based Nanoionic Resistive Memory
• A Configurable Memristor-based Finite Impulse Response Filter
• A light-stimulated neuromorphic device based on graphene hybrid phototransistor
• Markov chain Hebbian learning algorithm with ternary synaptic units
• Capacity, Fidelity, and Noise Tolerance of Associative Spatial-Temporal Memories Based on Memristive Neuromorphic Network
• Deep Neural Network Optimized to Resistive Memory with Nonlinear Current-Voltage Characteristics
• Practical Demonstration of a Memristive Fuse
• Deterministic Role of Concentration Surplus of Cation Vacancy over Anion Vacancy in Bipolar Memristive NiO
• Fast learning synapses with molecular spin valves via selective magnetic potentiation
• A Survey of Brain Inspired Technologies for Engineering