Peter Marwedel

Rainer Leupers

It is well known that embedded systems have to be implemented efficiently. This requires that processors optimized for certain application domains are used in embedded systems. Such an optimization requires a careful exploration of the design space, including a detailed study of cost/performance tradeoffs. In order to avoid time-consuming assembly language programming during design space exploration, compilers are needed. In order to analyze the effect of various software or hardware configurations on the performance, retargetable compilers are needed that can generate code for numerous different potential hardware configurations. This book provides a comprehensive and up-to-date overview of the fast developing area of retargetable compilers for embedded systems. It describes a large set important tools as well as applications of retargetable compilers at different levels in the design flow. Retargetable Compiler Technology for Embedded Systems is mostly self-contained and requires only fundamental knowledge in software and compiler design. It is intended to be a key reference for researchers and designers working on software, compilers, and processor optimization for embedded systems.

Tools and Applications

Retargetable Compiler Technology for Embedded Systems

Katharina Morik

Andreas Becker

Jens Buß

Machine learning is part of Artificial Intelligence since its beginning. Certainly, not learning would only allow the perfect being to show intelligent behavior. All others, be it humans or machines, need to learn in order to enhance their capabilities. In the eighties of the last century, learning from examples and modeling human learning strategies have been investigated in concert. The formal statistical basis of many learning methods has been put forward later on and is still an integral part of machine learning. Neural networks have always been in the toolbox of methods. Integrating all the pre-processing, exploitation of kernel functions, and transformation steps of a machine learning process into the architecture of a deep neural network increased the performance of this model type considerably. Modern machine learning is challenged on the one hand by the amount of data and on the other hand by the demand of real-time inference. This leads to an interest in computing architectures and modern processors. For a long time, the machine learning research could take the von-Neumann architecture for granted. All algorithms were designed for the classical CPU. Issues of implementation on a particular architecture have been ignored. This is no longer possible. The time for independently investigating machine learning and computational architecture is over. Computing architecture has experienced a similarly rampant development from mainframe or personal computers in the last century to now very large compute clusters on the one hand and ubiquitous computing of embedded systems in the Internet of Things on the other hand. Cyber-physical systems’ sensors produce a huge amount of streaming data which need to be stored and analyzed. Their actuators need to react in real-time. This clearly establishes a close connection with machine learning. Cyber-physical systems and systems in the Internet of Things consist of diverse components, heterogeneous both in hard- and software. Modern multi-core systems, graphic processors, memory technologies and hardware-software codesign offer opportunities for better implementations of machine learning models. Machine learning and embedded systems together now form a field of research which tackles leading edge problems in machine learning, algorithm engineering, and embedded systems. Machine learning today needs to make the resource demands of learning and inference meet the resource constraints of used computer architecture and platforms. A large variety of algorithms for the same learning method and, moreover, diverse implementations of an algorithm for particular computing architectures optimize learning with respect to resource efficiency while keeping some guarantees of accuracy. The trade-off between a decreased energy consumption and an increased error rate, to just give an example, needs to be theoretically shown for training a model and the model inference. Pruning and quantization are ways of reducing the resource requirements by either compressing or approximating the model. In addition to memory and energy consumption, timeliness is an important issue, since many embedded systems are integrated into large products that interact with the physical world. If the results are delivered too late, they may have become useless. As a result, real-time guarantees are needed for such systems. To efficiently utilize the available resources, e.g., processing power, memory, and accelerators, with respect to response time, energy consumption, and power dissipation, different scheduling algorithms and resource management strategies need to be developed. This book series addresses machine learning under resource constraints as well as the application of the described methods in various domains of science and engineering. Turning big data into smart data requires many steps of data analysis: methods for extracting and selecting features, filtering and cleaning the data, joining heterogeneous sources, aggregating the data, and learning predictions need to scale up. The algorithms are challenged on the one hand by high-throughput data, gigantic data sets like in astrophysics, on the other hand by high dimensions like in genetic data. Resource constraints are given by the relation between the demands for processing the data and the capacity of the computing machinery. The resources are runtime, memory, communication, and energy. Novel machine learning algorithms are optimized with regard to minimal resource consumption. Moreover, learned predictions are applied to program executions in order to save resources. The three books will have the following subtopics: Volume 1: Machine Learning under Resource Constraints - Fundamentals Volume 2: Machine Learning and Physics under Resource Constraints - Discovery Volume 3: Machine Learning under Resource Constraints - Applications Volume 1 establishes the foundations of this new field (Machine Learning under Resource Constraints). It goes through all the steps from data collection, their summary and clustering, to the different aspects of resource-aware learning, i.e., hardware, memory, energy, and communication awareness. Several machine learning methods are inspected with respect to their resource requirements and how to enhance their scalability on diverse computing architectures ranging from embedded systems to large computing clusters.

Fundamentals

Alberto Sangiovanni-Vincentelli

Marco Di Natale

Haibo Zeng

This book offers readers broad coverage of techniques to model, verify and validate the behavior and performance of complex distributed embedded systems. The authors attempt to bridge the gap between the three disciplines of model-based design, real-time analysis and model-driven development, for a better understanding of the ways in which new development flows can be constructed, going from system-level modeling to the correct and predictable generation of a distributed implementation, leveraging current and future research results.

From Functional Models to Implementations

Embedded Systems Development

Manish Verma

This book proposes novel memory hierarchies and software optimization techniques for the optimal utilization of memory hierarchies. It presents a wide range of optimizations, progressively increasing in the complexity of analysis and of memory hierarchies. The final chapter covers optimization techniques for applications consisting of multiple processes found in most modern embedded devices.

Advanced Memory Optimization Techniques for Low-Power Embedded Processors

Until the late 1980s, information processing was associated with large mainframe computers and huge tape drives. During the 1990s, this trend shifted toward information processing with personal computers, or PCs. The trend toward miniaturization continues and in the future the majority of information processing systems will be small mobile computers, many of which will be embedded into larger products and interfaced to the physical environment. Hence, these kinds of systems are called embedded systems. Embedded systems together with their physical environment are called cyber-physical systems. Examples include systems such as transportation and fabrication equipment. It is expected that the total market volume of embedded systems will be significantly larger than that of traditional information processing systems such as PCs and mainframes. Embedded systems share a number of common characteristics. For example, they must be dependable, efficient, meet real-time constraints and require customized user interfaces (instead of generic keyboard and mouse interfaces). Therefore, it makes sense to consider common principles of embedded system design. Embedded System Design starts with an introduction into the area and a survey of specification models and languages for embedded and cyber-physical systems. It provides a brief overview of hardware devices used for such systems and presents the essentials of system software for embedded systems, like real-time operating systems. The book also discusses evaluation and validation techniques for embedded systems. Furthermore, the book presents an overview of techniques for mapping applications to execution platforms. Due to the importance of resource efficiency, the book also contains a selected set of optimization techniques for embedded systems, including special compilation techniques. The book closes with a brief survey on testing. Embedded System Design can be used as a text book for courses on embedded systems and as a source which provides pointers to relevant material in the area for PhD students and teachers. It assumes a basic knowledge of information processing hardware and software. Courseware related to this book is available at http://ls12-www.cs.tu-dortmund.de/~marwedel.

Embedded Systems Foundations of Cyber-Physical Systems

Embedded System Design

Paul Lokuciejewski

For real-time systems, the worst-case execution time (WCET) is the key objective to be considered. Traditionally, code for real-time systems is generated without taking this objective into account and the WCET is computed only after code generation. Worst-Case Execution Time Aware Compilation Techniques for Real-Time Systems presents the first comprehensive approach integrating WCET considerations into the code generation process. Based on the proposed reconciliation between a compiler and a timing analyzer, a wide range of novel optimization techniques is provided. Among others, the techniques cover source code and assembly level optimizations, exploit machine learning techniques and address the design of modern systems that have to meet multiple objectives. Using these optimizations, the WCET of real-time applications can be reduced by about 30% to 45% on the average. This opens opportunities for decreasing clock speeds, costs and energy consumption of embedded processors. The proposed techniques can be used for all types real-time systems, including automotive and avionics IT systems.

Peter Marwedel

Books

Your next favorite book is out there