Resource-aware parameter tuning for real-time applications
Gabriel D, Stechele W, Wildermann S (2019)
Publication Type: Conference contribution
Publication year: 2019
Journal
Publisher: Springer Verlag
Book Volume: 11479 LNCS
Pages Range: 45-55
Conference Proceedings Title: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Event location: Copenhagen
ISBN: 9783030186555
DOI: 10.1007/978-3-030-18656-2_4
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.
Authors with CRIS profile
Involved external institutions
Germany (DE)How to cite
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: Download