Machine Learning and Non-volatile Memories

• Year: 2022
• Tongue: English
• Leaves: 178
• Category: Library

✦ Synopsis

Machine Learning and Non-volatile Memories (2022) [Micheloni Zambelli] [9783031038402].pdf

✦ Table of Contents

Foreword
Acknowledgments
From the Editors
Introduction
Contents
Editors and Contributors
Introduction to Machine Learning
1 Overview
2 Features
3 Learning Tasks
3.1 Supervised Learning of Predictive Models
4 Tree Models
4.1 Decision Trees
4.2 Random Forests
5 Linear Models
5.1 Training SVMs
5.2 Classification
5.3 Regression
6 Conclusions
References
Neural Networks and Deep Learning Fundamentals
1 A Brief History
2 Multilayer Perceptrons
3 Convolutional Neural Networks
4 Recurrent Neural Networks
5 Hyper-parameter Optimization
6 Conclusions
References
Accelerating Deep Neural Networks with Phase-Change Memory Devices
1 Introduction
1.1 The Promise of Analog AI
1.2 Two Common Neural Network Architectures
2 Software-Equivalent Accuracy in Analog AI
2.1 Programming Strategies
2.2 Counteracting Conductance Drift Using Slope Correction
3 Conclusion
References
Analogue In-Memory Computing with Resistive Switching Memories
1 Introduction
2 Resistive Switching Memories
2.1 Memory Array Structures
2.2 Requirements for Analogue Memory
3 In-Memory Computing Architecture for Matrix-Vector Multiplication
3.1 In-Memory Neural Network Accelerators
3.2 In-Memory Optimization Accelerators
4 In-Memory Computing Architecture for Inverse MVM
4.1 In-Memory Eigenvector Calculation
4.2 Pseudoinverse and Regression Accelerators
5 Conclusions
References
Introduction to 3D NAND Flash Memories
1 3D Charge Trap NAND Flash Memories
2 3D Floating Gate NAND Flash Memories
3 Key Challenges for 3D Flash Development
3.1 Number of Layers
3.2 Peripheral Circuits Under Memory Arrays
4 Future Trend for 3D NAND Flash
References
Deep Neural Network Engines Based on Flash Technology
1 Introduction
2 Deep Neural Networks Based on NOR Flash Memories
3 Deep Neural Networks Based on NAND Flash Memories
3.1 Conventional 3D NAND
3.2 3D NAND with Source-Line Read
3.3 3D NAND with Independent Source Lines
References
Machine Learning for 3D NAND Flash and Solid State Drives Reliability/Performance Optimization
1 Introduction
2 Reliability and Variability Issues of 3D NAND Flash
3 Machine Learning Techniques for 3D NAND Flash
3.1 Unsupervised Learning (Clustering) for ECC Optimization
3.2 Supervised Learning for Pre-emptive Endurance Prediction
3.3 Artificial Neural Networks Applied to Adaptive-ECC
3.4 Artificial Neural Networks for Variability Modeling
3.5 Recurrent Neural Networks for ECC Soft-Decision Speed Up
4 Machine Learning for Solid State Drives Prognostics
4.1 Machine Learning Assisted SSD Lifetime Enhancement
4.2 Echo State Networks for Hot Data Prediction in SSDs
5 Computational Storage and Machine Learning
5.1 The Computational Approximate Storage Concept
5.2 A Neural Network Engine Example in Computational Storage
6 Conclusions
References
Index