Название: Machine Learning Applications in Electronic Design Automation
Автор: Haoxing Ren, Jiang Hu
Издательство: Springer
Год: 2022
Страниц: 585
Язык: английский
Формат: pdf (true)
Размер: 19.1 MB

Electronic design automation (EDA) is a software technology that attempts to let computers undertake chip design tasks so that we can handle complexities beyond manual design capabilities. Although conventional EDA techniques have led to huge design productivity improvement, they face the fundamental limit that most EDA problems are NP hard and therefore have no polynomial-time algorithms for optimal solutions. As chip complexity increases to dozens of billions of transistors, such limitation becomes even more pronounced and there is a compelling need to have innovative changes.

This book serves as a single-source reference to key Machine Learning (ML) applications and methods in digital and analog design and verification. Experts from academia and industry cover a wide range of the latest research on ML applications in electronic design automation (EDA), including analysis and optimization of digital design, analysis and optimization of analog design, as well as functional verification, FPGA and system level designs, design for manufacturing (DFM), and design space exploration. The authors also cover key ML methods such as classical ML, Deep Learning models such as convolutional neural networks (CNNs), graph neural networks (GNNs), generative adversarial networks (GANs) and optimization methods such as reinforcement learning (RL) and Bayesian optimization (BO). All of these topics are valuable to chip designers and EDA developers and researchers working in digital and analog designs and verification.

Serves as a single-source reference to key Machine Learning (ML) applications and methods in digital and analog design and verification;
Covers classical ML methods, as well as Deep Learning models such as convolutional neural networks (CNNs), graph neural networks (GNNs), generative adversarial networks (GANs) and optimization methods such as reinforcement learning (RL) and Bayesian optimization (BO);
Discusses machine learning ML’s applications in electronic design automation (EDA), especially in the design automation of VLSI integrated circuits.

Contents:
Part I. Machine Learning-Based Design Prediction Techniques
1. ML for Design QoR Prediction
2. Deep Learning for Routability
3. Net-Based Machine Learning-Aided Approaches for Timing and Crosstalk Prediction
4. Deep Learning for Power and Switching Activity Estimation
5. Deep Learning for Analyzing Power Delivery Networks and Thermal Networks
6. Machine Learning for Testability Prediction
Part II. Machine Learning-Based Design Optimization Techniques
7. Machine Learning for Logic Synthesis
8. RL for Placement and Partitioning
9. Deep Learning Framework for Placement
10. Circuit Optimization for 2D and 3D ICs with Machine Learning
11. Reinforcement Learning for Routing
12. Machine Learning for Analog Circuit Sizing
Part III. Machine Learning Applications in Various Design Domains
13. The Interplay of Online and Offline Machine Learning for Design Flow Tuning
14. Machine Learning in the Service of Hardware Functional Verification
15. Machine Learning for Mask Synthesis and Verification
16. Machine Learning for Agile FPGA Design
17. Machine Learning for Analog Layout
18. ML for System-Level Modeling