Skip to main content
main-content

Über dieses Buch

This book discusses analysis, design and optimization techniques for streaming multiprocessor systems, while satisfying a given area, performance, and energy budget. The authors describe design flows for both application-specific and general purpose streaming systems. Coverage also includes the use of machine learning for thermal optimization at run-time, when an application is being executed. The design flow described in this book extends to thermal and energy optimization with multiple applications running sequentially and concurrently.

Inhaltsverzeichnis

Frontmatter

Chapter 1. Introduction

Multiprocessor systems have evolved over the past decades, triggered by innovations in transistor scaling and integration, multiprocessor design and system integration. This section summarizes these trends.
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 2. Operational Semantics of Application and Reliability Model

Synchronous Data Flow Graphs (SDFGs, see [12]) are often used for modeling modern DSP applications [17] and for designing concurrent multimedia applications implemented on multiprocessor systems. Both pipelined streaming and cyclic dependencies between tasks can be easily modeled in SDFGs. SDFGs allow analysis of a system in terms of throughput and other performance properties e.g., latency and buffer requirements [18]. Nodes of an SDFG are called actors; they represent functions that are computed by reading tokens (data items) from their input ports and writing results of computation as tokens on output ports. The number of tokens produced or consumed in one execution of an actor is called rate, and remains constant. Rates are visualized as port annotations. Actor execution is also called firing, and requires a fixed amount of time, denoted with a number in the actors. Edges in the graph, called channels, represent dependencies among different actors.
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 3. Literature Survey on System-Level Optimizations Techniques

As discussed in Chapter 1, design-time methodologies address three aspects – reliability and energy-aware platform-based design, reliability and energy-aware hardware-software co-design and energy-aware mapping for proactive fault-tolerance. Existing studies on these three aspects are discussed next.
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 4. Reliability and Energy-Aware Platform-Based Multiprocessor Design

As discussed in Chapter 3, a significant research is conducted recently to investigate platform-based design approaches in order to mitigate wear-out and minimize energy consumption. These studies, however, suffer from two limitations:
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 5. Reliability and Energy-Aware Co-design of Multiprocessor Systems

An emerging trend in multiprocessor design is to integrate reconfigurable area alongside homogeneous processing cores. Hardware-software co-design of these reconfigurable multiprocessor systems needs to address the following two aspects:
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 6. Design-Time Analysis for Fault-Tolerance

This chapter solves the following problem. Given a heterogeneous multiprocessor system and a set of multimedia applications, how to assign and order tasks of every application on the component cores such that the total energy consumption is minimized while guaranteeing to satisfy performance requirements of these application under all possible fault-scenarios. The scope of this work is limited to permanent failures of processing cores. Following are the key contributions of this chapter:
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 7. Run-Time Adaptations for Lifetime Improvement

As discussed in previous chapters, energy consumption and reliability are two important optimization objectives for multiprocessor systems. There is a strong interplay between these two objective functions. Reducing the temperature of a system by efficient thermal management leads to a reduction of leakage power. On the other hand, reduction of power dissipation (by controlling the voltage and frequency of operation) leads to an improvement in the thermal profile of a system. However, too frequent voltage and frequency scaling increases thermal cycles, which increases stress in the metal layers causing reliability concerns. This has attracted significant attention in recent years to investigate on intelligent techniques, such as the use of machine learning, to determine the relationship between temperature, energy and performance and their control using voltage and frequency switching. State-of-the-art learning-based approaches suffer from the following limitations.
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor

Chapter 8. Conclusions and Future Directions

Reliability and energy are emerging as two of the growing concerns for multiprocessor systems at deep sub-micron technology nodes. This book presented a system-level approach, namely application mapping and scheduling, to jointly address the reliability and energy problems for multiprocessor systems. A detailed background is presented in Chap. 3 on synchronous data flow graphs (SDFGs), used as application models in this book. This chapter also presented the mathematical background on lifetime reliability and the related studies on task mapping and scheduling for lifetime improvement.
Anup Kumar Das, Akash Kumar, Bharadwaj Veeravalli, Francky Catthoor
Weitere Informationen

BranchenIndex Online

Die B2B-Firmensuche für Industrie und Wirtschaft: Kostenfrei in Firmenprofilen nach Lieferanten, Herstellern, Dienstleistern und Händlern recherchieren.

Whitepaper

- ANZEIGE -

Globales Erdungssystem in urbanen Kabelnetzen

Bedingt durch die Altersstruktur vieler Kabelverteilnetze mit der damit verbundenen verminderten Isolationsfestigkeit oder durch fortschreitenden Kabelausbau ist es immer häufiger erforderlich, anstelle der Resonanz-Sternpunktserdung alternative Konzepte für die Sternpunktsbehandlung umzusetzen. Die damit verbundenen Fehlerortungskonzepte bzw. die Erhöhung der Restströme im Erdschlussfall führen jedoch aufgrund der hohen Fehlerströme zu neuen Anforderungen an die Erdungs- und Fehlerstromrückleitungs-Systeme. Lesen Sie hier über die Auswirkung von leitfähigen Strukturen auf die Stromaufteilung sowie die Potentialverhältnisse in urbanen Kabelnetzen bei stromstarken Erdschlüssen. Jetzt gratis downloaden!

Bildnachweise