Memristor Computing Systems

Description

This contributed volume offers practical solutions and design-, modeling-, and implementation-related insights that address current research problems in memristors, memristive devices, and memristor computing. The book studies and addresses related challenges in and proposes solutions for the future of memristor computing. State-of-the-art research on memristor modeling, memristive interconnections, memory circuit architectures,  software simulation tools, and applications of memristors in computing are presented. Utilising contributions from numerous experts in the field, written in clear language and illustrated throughout, this book is a comprehensive reference work. Memristor Computing Systems explains memristors and memristive devices in an accessible way for graduate students and researchers with a basic knowledge of electrical and control systems engineering, as well as prompting further research for more experienced academics. Preface Introduction Contents Part I Memristor Computing Theory 1 Edge-of-Chaos in CNN Models with Memristor Synapses 1.1 Introduction 1.2 Nonlinear Convection–Diffusion Model 1.2.1 Memristor CNN Model of (1.1) 1.2.2 Edge-of-Chaos in the M-CNN Model 1.2.3 Applications of the M-CNN Model to Noise Removal 1.3 M-CNN Model of Nano-structures 1.3.1 Statement of the Problem 1.3.2 M-CNN Model of Boundary Value Problem (1.8), (1.9) 1.3.3 Edge-of-Chaos in M-CNN Model (1.12) 1.3.4 Simulations and Validation 1.4 Conclusions References 2 Dynamic Analysis of Memristor Circuits via Input–Output Techniques 2.1 Introduction 2.2 Class of Circuits Description and Preliminaries 2.2.1 The Chua's Memristor Circuit 2.2.2 The Murali–Lakshmanan–Chua's Memristor Circuit 2.3 A Canonical Reduced-Order Input–Output Representation 2.3.1 Canonical Input–Output Representation of Chua's Memristor Circuit 2.3.2 Canonical Input–Output Representation of Murali–Lakshmanan–Chua's Memristor Circuit 2.4 Circuit Implementation of the Canonical Input–Output Representation 2.4.1 The Case of Chua's Memristor Circuit 2.4.2 The Case of Murali–Lakshmanan–Chua's Memristor Circuit 2.5 Numerical Examples 2.5.1 Chua's Memristor Circuit 2.5.2 Murali–Lakshmanan–Chua's Memristor Circuit 2.6 Conclusion References 3 Energy-Based Memristor Networks for Pattern Recognition in Vision Systems 3.1 Introduction 3.2 Energy-Based Neural Networks 3.2.1 Equilibrium Propagation 3.3 Memristor-Based Recurrent Neural Network 3.4 Simulations 3.5 Conclusions References 4 Tunable Chaos in Memristor Circuits for Pattern Recognition Tasks 4.1 Introduction 4.2 Flux-Charge Analysis Method 4.3 The Memristor-Based Chua's Oscillator 4.4 Bio-inspired Analog Computing with a Source of Tunable Chaos 4.4.1 The Architecture 4.4.2 The Training 4.4.3 Simulation Methodology 4.4.4 Simulation Results 4.5 Conclusions References 5 Pattern Formation in an M-CNN Structure Utilizing a Locally Active NbOx Memristor 5.1 Introduction 5.2 The Locally Active NbOx Memristor 5.2.1 The Model Equations 5.2.2 The AC Equivalent Circuit 5.3 The Single Cell 5.3.1 Stability Analysis 5.3.2 Local Activity Analysis 5.3.3 Parameter Space Analysis 5.4 M-CNN Structure and Pattern Formation 5.5 Conclusion References Part II In-Memory Computing 6 In-Memory Computing with Non-volatile Memristor CAM Circuits 6.1 Introduction 6.2 Content-Addressable Memory (CAM) Circuits 6.2.1 Memristor CAM Circuits 6.2.2 Analog Memristor CAM Circuits 6.2.3 CAM-Specific Operation Advantages 6.3 Computing with CAMs 6.3.1 Finite Automata 6.3.2 Associative Computing Systems 6.3.3 Locality Sensitive Hashing 6.3.4 Tree-Based Models 6.4 Conclusion and Outlook References 7 Memristor-Based In-Memory Computing Architecture for Scientific Computing 7.1 Introduction 7.2 Vector–Matrix Multiplication in Memristor Crossbar Array 7.3 In-Memory Scientific Computing Acceleration 7.3.1 Solving Systems of Linear Equations 7.3.2 Solving Partial Differential Equations 7.4 Discussion and Outlook 7.4.1 Device Level 7.4.2 System Level 7.4.3 Conclusion References 8 Ta/HfO2-based Memristor and Crossbar Arrays for In-Memory Computing 8.1 Introduction 8.2 Highly Reliable Ta/HfO2 Memristors 8.2.1 Structure, Fabrication, and Electrical Behavior 8.2.2 Switching Mechanism Studies 8.3 1T1R Array: Integration and Operation 8.4 In-Memory Computing for Neural Networks 8.4.1 Multi-layer Perceptron 8.4.2 Convolutional Neural Networks 8.4.3 Long Short-Term Memory Networks 8.4.4 Reinforcement Learning 8.5 In-Memory Computing with Embedded Security 8.6 Conclusions References Part III Memristive Devices 9 Ionic Nanoarchitectonics for Artificial Intelligence Devices 9.1 Introduction 9.2 Synaptic Behavior of Gap-Type and Gapless-Type Atomic Switches 9.3 Advanced Gap-Type Atomic Switches Using Molecular Layers 9.4 Ionic Decision-Maker for Reinforcement Learning Using Electromotive Force 9.5 Summary References 10 Optical Memristors: Review of Switching Mechanisms and New Computing Paradigms 10.1 Introduction 10.2 Memristor Basics 10.3 Optical Memristors and Switching Mechanisms 10.3.1 Barrier Modification, Photogating and Photoconductance Effects 10.3.2 Chemical-Based Approaches 10.3.3 Plasmonic Memristors 10.4 Optical Computing Paradigms and Optical Memristor Applications 10.4.1 Optical Memristor Cellular Nonlinear Networks (OM-CNN) for Ultra-Fast Vision Recognition Applications 10.4.2 Neuromorphic Computing: Light Tuneable STDP Learning 10.5 Conclusion and Outlook References Part IV Bioinspired Memristor Computing 11 Memristive Models for the Emulation of Biological Learning 11.1 Introduction 11.2 Learning Principles in Biology 11.2.1 Synaptic Plasticity 11.2.2 Long-Term Potentiation 11.2.3 Hebb's Learning Rule and Spike Time-Dependent Plasticity 11.3 Memristive Learning Models 11.3.1 Hebbian Learning 11.3.2 Memristive Hebbian Learning Models 11.4 Conclusion and Outlook References 12 Organic Memristive Devices and Organic Electrochemical Transistors as Promising Elements for Bio-inspired Systems 12.1 The Meaning and the Importance of Bio-inspired Systems 12.1.1 Bio-inspired Computing 12.1.2 Organic Devices in Neuromorphic Computing 12.2 OMD Bio-inspired Applications 12.2.1 OMDs: Structure and Working Principle 12.2.2 OMDs as Logic Gates 12.2.3 OMDs as Synapse Analogues 12.2.4 Networks 12.2.5 Interfacing Real Neurons 12.3 OECT Bio-inspired Applications 12.3.1 OECTs: Structure and Working Mechanism 12.3.2 OECTs in Neuromorphic Applications 12.4 Conclusions and Perspectives References Index