Evaluating System Identification Methods for Predicting Thermal Dissipation of Heterogeneous SoCs

Research output: Chapter in Book/Conference proceedingConference contributionScientificpeer-review

8 Downloads (Pure)

Abstract

In this paper we evaluate the use of system identification methods to build a thermal prediction model of heterogeneous SoC platforms that can be used to quickly predict the temperature of difierent configurations without the need of hardware. Specifically, we focus on modeling approaches that can predict the temperature based on the clock frequency and the utilization percentage of each core.
We investigate three methods with respect to their prediction accuracy: a linear state-space identification approach using polynomial regressors, a NARX neural network approach and a recurrent neural network approach configured in an FIR model structure. We evaluate the methods on an Odroid-XU4 board featuring an Exynos 5422 SoC. The results show that the model based on polynomial regressors significantly outperformed the other two models when trained with 1 hour and 6 hours of data.
Original languageEnglish
Title of host publicationEmbedded Computer Systems: Architectures, Modeling, and Simulation. SAMOS 2021.
PublisherSpringer
Pages144-160
Number of pages17
ISBN (Print)978-3-031-04579-0
DOIs
Publication statusPublished - 2022
MoE publication typeA4 Article in a conference publication
Event21st International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation, SAMOS 2021 - Virtual, Online
Duration: 4 Jul 20218 Jul 2021

Publication series

Name Lecture Notes in Computer Science
PublisherSpringer
Volume13227
ISSN (Print)0302-9743

Conference

Conference21st International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation, SAMOS 2021
CityVirtual, Online
Period04/07/2108/07/21

Keywords

  • energy dissipation
  • Temperature
  • Multicore processing
  • System identification

Fingerprint

Dive into the research topics of 'Evaluating System Identification Methods for Predicting Thermal Dissipation of Heterogeneous SoCs'. Together they form a unique fingerprint.

Cite this