• English
  • Deutsch
  • Log In
    Password Login
    Research Outputs
    Fundings & Projects
    Researchers
    Institutes
    Statistics
Repository logo
Fraunhofer-Gesellschaft
  1. Home
  2. Fraunhofer-Gesellschaft
  3. Konferenzschrift
  4. Partitioning and Task Transfer on NoC-based Many-Core Processors in the Avionics Domain
 
  • Details
  • Full
Options
2011
Conference Paper
Title

Partitioning and Task Transfer on NoC-based Many-Core Processors in the Avionics Domain

Abstract
Networks-on-Chip (NoC) based many-core processors can not only increase system performance but also allow the integration of multiple functions on a single hardware platform. To consolidate functionality on many-core systems in safety-critical domains software partitioning is required to avoid the propagation of faults due to the use of shared resources. In this paper we propose extensions to well established single-core partitioning mechanisms to take novel architectural characteristics of many-core processors into account. In addition to fixed partitioning, we present flexible partitioning as an approach to improve resource utilization and fault tolerance using dynamic reconfiguration. Flexible partitioning requires task migration between cores via a shared resource - the N oC - which may endanger the required predictability. Therefore we empirically analyze a variety of task transfer mechanisms on a Tilera TILEPro64TM many-core processor regarding their potential for deterministic reconfiguration during run-time.
Author(s)
Hilbrich, Robert
Kampenhout, J. Reinier van
Mainwork
Berichte aus den Fachgruppen und Arbeitskreisen : FG Ada4. Workshop Entwicklung Zuverlässiger Software-Systeme 2011. Bericht und Beiträge  
Conference
Workshop Entwicklung zuverlässiger Software-Systeme 2011  
File(s)
Download (246.11 KB)
Rights
Use according to copyright law
DOI
10.24406/publica-fhg-375234
Language
German
FIRST
Keyword(s)
  • Partitioning

  • Task Transfer

  • Task Migration

  • multicore

  • NoC

  • Cookie settings
  • Imprint
  • Privacy policy
  • Api
  • Contact
© 2024