Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware

Gaurav Mitra; Andrew Haigh; Anish Varghese; Luke Angove; Alistair P. Rendell

doi:10.1109/HPCC-SmartCity-DSS.2016.0113

Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware

Gaurav Mitra
, Andrew Haigh
, Anish Varghese
, Luke Angove
, Alistair P. Rendell

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

Heterogeneous System-on-Chip (SoC) processors are increasingly gaining traction in the High Performance Computing (HPC) community as alternate building blocks for future exascale systems. Key issues relating to their promise of energy efficiency include i) absolute performance, ii) finding an energy-optimal balance in the use of different on-chip devices and iii) understanding the performance-energy trade-offs while using different on-chip devices. In this paper we explore these issues through an energy usage model designed to predict the existence of an energy-optimal work partition between different processing elements on heterogeneous systems for any application. We validate our model by measuring performance and energy consumption of matrix multiplication on the NVIDIA Tegra K1 and X1 systems. An environment for monitoring and responding to energy usage is also outlined and used to perform high resolution measurements. Comparisons are drawn with conventional HPC systems housing Intel Xeon CPUs alongside NVIDIA GPUs.

Original language	English
Title of host publication	Proceedings
Subtitle of host publication	18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016
Editors	Laurence T. Yang, Jinjun Chen
Publisher	Institute of Electrical and Electronics Engineers
Pages	781-788
Number of pages	8
ISBN (Electronic)	9781509042968
DOIs	https://doi.org/10.1109/HPCC-SmartCity-DSS.2016.0113
Publication status	Published - 20 Jan 2017
Externally published	Yes
Event	18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016 - Sydney, Australia Duration: 12 Dec 2016 → 14 Dec 2016

Publication series

Name	Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016

Conference

Conference	18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016
Country/Territory	Australia
City	Sydney
Period	12/12/16 → 14/12/16

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

Energy efficiency
Energy usage model
Haswell
Intel
K20
K80
Load balancing
NVIDIA
Power Measurement
Sandy bridge
SoC
Tegra K1
Tegra X1
UCurrent

Access to Document

10.1109/HPCC-SmartCity-DSS.2016.0113Licence: In Copyright

Cite this

Mitra, G., Haigh, A., Varghese, A., Angove, L., & Rendell, A. P. (2017). Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware. In L. T. Yang, & J. Chen (Eds.), Proceedings: 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016 (pp. 781-788). Article 7828454 (Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016). Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/HPCC-SmartCity-DSS.2016.0113

Mitra, Gaurav ; Haigh, Andrew ; Varghese, Anish et al. / Split Wisely : When Work Partitioning is Energy-Optimal on Heterogeneous Hardware. Proceedings: 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016. editor / Laurence T. Yang ; Jinjun Chen. Institute of Electrical and Electronics Engineers, 2017. pp. 781-788 (Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016).

@inproceedings{276e731cc71b41b4a3001a39eaca674c,

title = "Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware",

abstract = "Heterogeneous System-on-Chip (SoC) processors are increasingly gaining traction in the High Performance Computing (HPC) community as alternate building blocks for future exascale systems. Key issues relating to their promise of energy efficiency include i) absolute performance, ii) finding an energy-optimal balance in the use of different on-chip devices and iii) understanding the performance-energy trade-offs while using different on-chip devices. In this paper we explore these issues through an energy usage model designed to predict the existence of an energy-optimal work partition between different processing elements on heterogeneous systems for any application. We validate our model by measuring performance and energy consumption of matrix multiplication on the NVIDIA Tegra K1 and X1 systems. An environment for monitoring and responding to energy usage is also outlined and used to perform high resolution measurements. Comparisons are drawn with conventional HPC systems housing Intel Xeon CPUs alongside NVIDIA GPUs.",

keywords = "Energy efficiency, Energy usage model, Haswell, Intel, K20, K80, Load balancing, NVIDIA, Power Measurement, Sandy bridge, SoC, Tegra K1, Tegra X1, UCurrent",

author = "Gaurav Mitra and Andrew Haigh and Anish Varghese and Luke Angove and Rendell, \{Alistair P.\}",

year = "2017",

month = jan,

day = "20",

doi = "10.1109/HPCC-SmartCity-DSS.2016.0113",

language = "English",

series = "Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016",

publisher = "Institute of Electrical and Electronics Engineers",

pages = "781--788",

editor = "Yang, \{Laurence T.\} and Jinjun Chen",

booktitle = "Proceedings",

address = "United States",

note = "18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016 ; Conference date: 12-12-2016 Through 14-12-2016",

}

Mitra, G, Haigh, A, Varghese, A, Angove, L & Rendell, AP 2017, Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware. in LT Yang & J Chen (eds), Proceedings: 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016., 7828454, Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016, Institute of Electrical and Electronics Engineers, pp. 781-788, 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016, Sydney, Australia, 12/12/16. https://doi.org/10.1109/HPCC-SmartCity-DSS.2016.0113

Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware. / Mitra, Gaurav; Haigh, Andrew; Varghese, Anish et al.
Proceedings: 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016. ed. / Laurence T. Yang; Jinjun Chen. Institute of Electrical and Electronics Engineers, 2017. p. 781-788 7828454 (Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - Split Wisely

T2 - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016

AU - Mitra, Gaurav

AU - Haigh, Andrew

AU - Varghese, Anish

AU - Angove, Luke

AU - Rendell, Alistair P.

PY - 2017/1/20

Y1 - 2017/1/20

N2 - Heterogeneous System-on-Chip (SoC) processors are increasingly gaining traction in the High Performance Computing (HPC) community as alternate building blocks for future exascale systems. Key issues relating to their promise of energy efficiency include i) absolute performance, ii) finding an energy-optimal balance in the use of different on-chip devices and iii) understanding the performance-energy trade-offs while using different on-chip devices. In this paper we explore these issues through an energy usage model designed to predict the existence of an energy-optimal work partition between different processing elements on heterogeneous systems for any application. We validate our model by measuring performance and energy consumption of matrix multiplication on the NVIDIA Tegra K1 and X1 systems. An environment for monitoring and responding to energy usage is also outlined and used to perform high resolution measurements. Comparisons are drawn with conventional HPC systems housing Intel Xeon CPUs alongside NVIDIA GPUs.

AB - Heterogeneous System-on-Chip (SoC) processors are increasingly gaining traction in the High Performance Computing (HPC) community as alternate building blocks for future exascale systems. Key issues relating to their promise of energy efficiency include i) absolute performance, ii) finding an energy-optimal balance in the use of different on-chip devices and iii) understanding the performance-energy trade-offs while using different on-chip devices. In this paper we explore these issues through an energy usage model designed to predict the existence of an energy-optimal work partition between different processing elements on heterogeneous systems for any application. We validate our model by measuring performance and energy consumption of matrix multiplication on the NVIDIA Tegra K1 and X1 systems. An environment for monitoring and responding to energy usage is also outlined and used to perform high resolution measurements. Comparisons are drawn with conventional HPC systems housing Intel Xeon CPUs alongside NVIDIA GPUs.

KW - Energy efficiency

KW - Energy usage model

KW - Haswell

KW - Intel

KW - K20

KW - K80

KW - Load balancing

KW - NVIDIA

KW - Power Measurement

KW - Sandy bridge

KW - SoC

KW - Tegra K1

KW - Tegra X1

KW - UCurrent

UR - https://www.scopus.com/pages/publications/85013674927

U2 - 10.1109/HPCC-SmartCity-DSS.2016.0113

DO - 10.1109/HPCC-SmartCity-DSS.2016.0113

M3 - Conference contribution

AN - SCOPUS:85013674927

T3 - Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016

SP - 781

EP - 788

BT - Proceedings

A2 - Yang, Laurence T.

A2 - Chen, Jinjun

PB - Institute of Electrical and Electronics Engineers

Y2 - 12 December 2016 through 14 December 2016

ER -

Mitra G, Haigh A, Varghese A, Angove L, Rendell AP. Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware. In Yang LT, Chen J, editors, Proceedings: 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016. Institute of Electrical and Electronics Engineers. 2017. p. 781-788. 7828454. (Proceedings - 18th IEEE International Conference on High Performance Computing and Communications, 14th IEEE International Conference on Smart City and 2nd IEEE International Conference on Data Science and Systems, HPCC/SmartCity/DSS 2016). doi: 10.1109/HPCC-SmartCity-DSS.2016.0113

Split Wisely: When Work Partitioning is Energy-Optimal on Heterogeneous Hardware

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this