Performance Evaluation of Thermal-Constrained Scheduling Strategies in Multi-core Systems
From International Center for Computational Logic
Performance Evaluation of Thermal-Constrained Scheduling Strategies in Multi-core Systems
Muhammad Usama SardarMuhammad Usama Sardar, Clemens DubslaffClemens Dubslaff, Sascha KlüppelholzSascha Klüppelholz, Christel BaierChristel Baier, Akash KumarAkash Kumar
Muhammad Usama Sardar, Clemens Dubslaff, Sascha Klüppelholz, Christel Baier, Akash Kumar
Performance Evaluation of Thermal-Constrained Scheduling Strategies in Multi-core Systems
Proc. of 16th European Workshop on Computer Performance Engineering (EPEW), volume 12039 of Lecture Notes in Computer Science, 133--147, 2019. Springer
Performance Evaluation of Thermal-Constrained Scheduling Strategies in Multi-core Systems
Proc. of 16th European Workshop on Computer Performance Engineering (EPEW), volume 12039 of Lecture Notes in Computer Science, 133--147, 2019. Springer
- KurzfassungAbstract
The increasing usage of multi-cores in safety-critical applications, such as autonomous control, demands high levels of reliability, which crucially depends on the temperature. On the other hand, there is a natural trade-off between reliability and performance. The scheduling of tasks is one of the key factors which determine the resulting system performance as well as reliability. Commonly used techniques, such as simulation based on benchmarks, can simulate only a limited number of input sequences of system runs and hardly optimize the performance-reliability trade-off. In order to accurately evaluate the schedulers and provide formal guarantees suitable in early design stages, we use formal methods for a quantitative performance-reliability trade-off analysis. Specifically, we propose to use energy-utility quantiles as a metric to evaluate the effectiveness of a given scheduler. For illustration, we evaluate TAPE, a state-of-the-art thermal-constrained scheduler, with theoretical optimal ones. - Forschungsgruppe:Research Group: Algebraische und logische Grundlagen der InformatikAlgebraic and Logical Foundations of Computer Science
@inproceedings{SDKBK2019,
author = {Muhammad Usama Sardar and Clemens Dubslaff and Sascha
Kl{\"{u}}ppelholz and Christel Baier and Akash Kumar},
title = {Performance Evaluation of Thermal-Constrained Scheduling
Strategies in Multi-core Systems},
booktitle = {Proc. of 16th European Workshop on Computer Performance
Engineering (EPEW)},
series = {Lecture Notes in Computer Science},
volume = {12039},
publisher = {Springer},
year = {2019},
pages = {133--147},
doi = {10.1007/978-3-030-44411-2_9}
}