Resource-aware parameter tuning for real-time applications

Beitrag bei einer Tagung


Details zur Publikation

Autorinnen und Autoren: Gabriel D, Stechele W, Wildermann S
Herausgeber: Martin Schoeberl, Thilo Pionteck, Sascha Uhrig, Jürgen Brehm, Christian Hochberger
Verlag: Springer Verlag
Jahr der Veröffentlichung: 2019
Band: 11479 LNCS
Tagungsband: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Seitenbereich: 45-55
ISBN: 9783030186555
ISSN: 0302-9743


Abstract

Executing multiple applications on a multi-core system while the workload of all applications varies brings the challenge of dynamically adapting resource allocations and parametrizing with respect to constraints e.g. timing limits of real-time applications. We present a hybrid approach which extracts a set of Pareto-optimal operating points during design time which are used to dynamically parameterize the periodic application during run-time. The setup is done at the beginning of each iteration of the execution and exclusively allocates processing elements from the system depending on the current workload. The parametrization is performed with the observed information about workload complexity and allocated resources. Therefore guarantees on time limits can be granted for all iterations including situations when the number of available processing elements has been decreased sharply.


FAU-Autorinnen und Autoren / FAU-Herausgeberinnen und Herausgeber

Wildermann, Stefan Dr.-Ing.
Lehrstuhl für Informatik 12 (Hardware-Software-Co-Design)


Einrichtungen weiterer Autorinnen und Autoren

Technische Universität München (TUM)


Zitierweisen

APA:
Gabriel, D., Stechele, W., & Wildermann, S. (2019). Resource-aware parameter tuning for real-time applications. In Martin Schoeberl, Thilo Pionteck, Sascha Uhrig, Jürgen Brehm, Christian Hochberger (Eds.), Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (pp. 45-55). Copenhagen, DK: Springer Verlag.

MLA:
Gabriel, Dirk, Walter Stechele, and Stefan Wildermann. "Resource-aware parameter tuning for real-time applications." Proceedings of the 32nd International Conference on Architecture of Computing Systems, ARCS 2019, Copenhagen Ed. Martin Schoeberl, Thilo Pionteck, Sascha Uhrig, Jürgen Brehm, Christian Hochberger, Springer Verlag, 2019. 45-55.

BibTeX: 

Zuletzt aktualisiert 2019-03-06 um 10:08