@inproceedings{0ed5646ce75445b2ac2980e1ffddbb6a,
title = "Performance and energy analysis of scientific workloads executing on LPSoCs",
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.",
keywords = "DVFS, Energy efficiency, Energy usage model, LPSoC, Tegra SoC",
author = "Anish Varghese and Joshua Milthorpe and Rendell, {Alistair P.}",
note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG, part of Springer Nature 2018.; 12th International Conference on Parallel Processing and Applied Mathematics, PPAM 2017 ; Conference date: 10-09-2017 Through 13-09-2017",
year = "2018",
doi = "10.1007/978-3-319-78054-2_11",
language = "English",
isbn = "9783319780535",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "113--122",
editor = "Ewa Deelman and Roman Wyrzykowski and Konrad Karczewski and Jack Dongarra",
booktitle = "Parallel Processing and Applied Mathematics - 12th International Conference, PPAM 2017, Revised Selected Papers",
address = "Germany",
}