Performance and energy analysis of scientific workloads executing on LPSoCs

Anish Varghese, Joshua Milthorpe, Alistair P. Rendell

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Low-power system-on-chip (LPSoC) processors provide an interesting alternative as building blocks for future HPC systems due to their high energy efficiency. However, understanding their performance-energy trade-offs and minimizing the energy-to-solution for an application running across the heterogeneous devices of an LPSoC remains a challenge. In this paper, we describe our methodology for developing an energy model which may be used to predict the energy usage of application code executing on an LPSoC system under different frequency settings. For this paper, we focus only on the CPU. Performance and energy measurements are presented for different types of workloads on the NVIDIA Tegra TK1 and Tegra TX1 systems at varying frequencies. From these results, we provide insights on how to develop a model to predict energy usage at different frequencies for general workloads.

Original languageEnglish
Title of host publicationParallel Processing and Applied Mathematics - 12th International Conference, PPAM 2017, Revised Selected Papers
EditorsEwa Deelman, Roman Wyrzykowski, Konrad Karczewski, Jack Dongarra
PublisherSpringer-Verlag
Pages113-122
Number of pages10
Volume10778 LNCS
ISBN (Print)9783319780535
DOIs
Publication statusPublished - 10 Sep 2017
Externally publishedYes
Event12th International Conference on Parallel Processing and Applied Mathematics, PPAM 2017 - Czestochowa, Poland
Duration: 10 Sep 201713 Sep 2017

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume10778 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference12th International Conference on Parallel Processing and Applied Mathematics, PPAM 2017
CountryPoland
CityCzestochowa
Period10/09/1713/09/17

Bibliographical note

© Springer International Publishing AG, part of Springer Nature 2018

Keywords

  • DVFS
  • Energy efficiency
  • Energy usage model
  • LPSoC
  • Tegra SoC

Fingerprint Dive into the research topics of 'Performance and energy analysis of scientific workloads executing on LPSoCs'. Together they form a unique fingerprint.

  • Cite this

    Varghese, A., Milthorpe, J., & Rendell, A. P. (2017). Performance and energy analysis of scientific workloads executing on LPSoCs. In E. Deelman, R. Wyrzykowski, K. Karczewski, & J. Dongarra (Eds.), Parallel Processing and Applied Mathematics - 12th International Conference, PPAM 2017, Revised Selected Papers (Vol. 10778 LNCS, pp. 113-122). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10778 LNCS). Springer-Verlag. https://doi.org/10.1007/978-3-319-78054-2_11