| Author | Francky Catthoor, Koen Danckaert, Chidamber Kulkarni, Erik Brockmeyer, Per Gunnar Kjeldsberg, Tanja Van Achteren, Thierry Omnes |
| Publisher | Springer |
| Year | 2002 |
| Language | English |
| Pages | 306 |
| Size | 4.7 MB |
| Extension |
Summary
Data Access and Storage Management for Embedded Programmable Processors, authored by a team of leading researchers from IMEC, provides a comprehensive overview of the state-of-the-art in system-level data access and storage management for embedded programmable processors[reference:0][reference:1]. The book targets the complex domain of embedded real-time multi-media and communication applications, which are characterized as "data-dominated" – meaning that cost-related aspects like power consumption and silicon footprint are heavily influenced, if not dominated, by data access and storage patterns[reference:2][reference:3].
The material presented in this volume is primarily based on extensive research conducted at IMEC between 1996 and 2001[reference:4][reference:5]. To address the stringent timing requirements and data-dominated nature of this application domain, the authors adopt a target architecture style compatible with modern embedded processors. They develop a systematic, step-wise methodology known as Data Transfer and Storage Exploration (DTSE), which makes the exploration and optimization of such applications feasible through a source-to-source precompilation approach[reference:6][reference:7].
The book is organized into three main parts. Part 1 introduces the context and motivation, providing a high-level overview of the entire DTSE approach, illustrated with a relevant driver from the targeted application domain[reference:8]. Part 2 demonstrates how source-to-source code transformations play a crucial role in solving the data transfer and storage bottleneck in modern processor architectures[reference:9]. These transformations, which include global data-flow and loop transformations, data reuse decisions, and high-level estimators, are shown to be platform-independent, resulting in optimized code that performs better on any modern platform[reference:10]. Part 3 discusses research efforts related to mapping embedded applications to specific memory organizations in embedded programmable processors[reference:11]. This section covers storage cycle budget distribution, support for modern memory architectures like SDRAMs, and cache-related issues, emphasizing the trade-off between timing characteristics and cost effects[reference:12].
The book is an essential resource for researchers, practicing engineers, and graduate students in the fields of embedded systems, compiler design, and electronic design automation. It offers a deep dive into the challenges and solutions for managing data access and storage in modern embedded processors, providing both theoretical foundations and practical methodologies.
Key Features
- Comprehensive Overview: Provides a complete picture of system-level data access and storage management for embedded programmable processors[reference:13].
- Focus on Data-Dominated Applications: Specifically addresses the challenges of real-time multi-media and communication applications where data access and storage dominate cost[reference:14].
- Systematic Methodology (DTSE): Presents a step-wise Data Transfer and Storage Exploration methodology for systematic optimization[reference:15].
- Source-to-Source Compilation: Emphasizes a precompilation approach to make exploration and optimization feasible[reference:16].
- Platform-Independent Transformations: Describes code transformations that are effective across different modern platforms[reference:17].
- Practical Techniques: Covers global loop transformations, data reuse exploration, high-level storage estimation, and cache optimization[reference:18].
- Real-World Demonstrators: Includes demonstrator designs to illustrate the application of the proposed methodologies[reference:19].
- Industry-Relevant Research: Based on cutting-edge research conducted at IMEC, a world-renowned research center[reference:20][reference:21].
- Addresses Key Cost Factors: Directly tackles power consumption and memory footprint, critical concerns in embedded system design[reference:22].
- Valuable Reference: Serves as an essential reference for researchers, engineers, and students in embedded systems and related fields.
About Author
Francky Catthoor is a prominent researcher in the field of embedded systems and data-dominated applications. He is a key figure at IMEC (Interuniversity Microelectronics Centre), where he has led numerous research projects on system-level design and optimization. His work focuses on data transfer and storage management, memory architecture, and power-aware design. He has authored and co-authored several influential books and papers in these areas.
Koen Danckaert, Chidamber Kulkarni, Erik Brockmeyer, Per Gunnar Kjeldsberg, Tanja Van Achteren, and Thierry Omnes are all esteemed researchers who have contributed significantly to the field of embedded systems. Their collaborative work at IMEC has led to groundbreaking advancements in system-level data access and storage management, compiler optimization, and memory architecture design[reference:23][reference:24]. Their combined expertise spans computer architecture, compiler technology, and electronic design automation, making this book a definitive reference in the domain.
Related Books
- Embedded System Design - Peter Marwedel
- Memory Architecture and Data Access Optimization for Embedded Systems - Francky Catthoor, et al.
- Compiler Construction: Principles and Practice - Kenneth C. Louden
- Advanced Compiler Design and Implementation - Steven Muchnick
- Power-Aware Computer Systems - Babak Falsafi, T. N. Vijaykumar
- Low Power Digital CMOS Design - Anantha P. Chandrakasan, Robert W. Brodersen
- Hardware/Software Co-Design: Principles and Practice - Jørgen Staunstrup, Wayne Wolf
- Real-Time Systems - Jane W. S. Liu
Ads
Frequently Asked Questions
Q: What is the main focus of this book?
A: The book focuses on system-level data access and storage management for embedded programmable processors, particularly for data-dominated real-time multi-media and communication applications[reference:25].
Q: Who are the authors of this book?
A: The book is authored by Francky Catthoor, Koen Danckaert, Chidamber Kulkarni, Erik Brockmeyer, Per Gunnar Kjeldsberg, Tanja Van Achteren, and Thierry Omnes[reference:26][reference:27].
Q: What is the Data Transfer and Storage Exploration (DTSE) methodology?
A: DTSE is a systematic step-wise methodology developed by the authors to make the exploration and optimization of data access and storage in embedded applications feasible through a source-to-source precompilation approach[reference:28].
Q: Is this book suitable for graduate students?
A: Yes, this book is an excellent resource for graduate students in computer engineering, electrical engineering, and computer science, particularly those specializing in embedded systems, compiler design, and electronic design automation.
Q: What are the key application domains covered in this book?
A: The book targets complex embedded real-time multi-media and communication applications, which are characterized by being data-dominated[reference:29].
Q: Does the book address power consumption and memory footprint?
A: Yes, the book extensively addresses these critical cost factors, which are heavily influenced by data access and storage aspects in embedded systems[reference:30].
Q: What is the publication date of this book?
A: The book was published in 2002[reference:31][reference:32].
Enregistrer un commentaire
Thanks for comment