Transitioning HPC software to exascale heterogeneous computing

Wen Mei Hwu; Li Wen Chang; Hee Seok Kim; Abdul Dakkak; Izzat El Hajj

doi:10.1109/CEM.2015.7237412

Transitioning HPC software to exascale heterogeneous computing

Wen Mei Hwu, Li Wen Chang, Hee Seok Kim, Abdul Dakkak, Izzat El Hajj

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

Abstract

An increasing portion of the top supercomputers in the world, including Blue Waters, have heterogeneous CPUGPU computational units. As we move towards exascale, we can expect even more pervasive deployment of heterogeneous computational units. While a handful of science teams can already use heterogeneous computational units in their production applications, there is still significant room for the growing use. This paper presents the current state and projected path for transitioning software into this new paradigm. We first summarize the currently practical languages such as OpenCL, OpenACC, and C++AMP, in increasing levels of productivity, highlighting their recent advancements in supporting performance portability and maintainability. We will then give a brief overview of some emerging programming systems such as TANGRAM and Troilet that are designed to further enhance developer productivity for heterogeneous computing.

Original language	English (US)
Title of host publication	2015 Computational Electromagnetics International Workshop, CEM 2015
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	4-5
Number of pages	2
ISBN (Electronic)	9781467371971
DOIs	https://doi.org/10.1109/CEM.2015.7237412
State	Published - Sep 2 2015
Event	Computational Electromagnetics International Workshop, CEM 2015 - Izmir, Turkey Duration: Jul 1 2015 → Jul 4 2015

Publication series

Name	2015 Computational Electromagnetics International Workshop, CEM 2015

Other

Other	Computational Electromagnetics International Workshop, CEM 2015
Country/Territory	Turkey
City	Izmir
Period	7/1/15 → 7/4/15

ASJC Scopus subject areas

Electrical and Electronic Engineering
Computational Mathematics

Online availability

10.1109/CEM.2015.7237412

Library availability

Discover UIUC Full Text

Cite this

Hwu, W. M., Chang, L. W., Kim, H. S., Dakkak, A., & El Hajj, I. (2015). Transitioning HPC software to exascale heterogeneous computing. In 2015 Computational Electromagnetics International Workshop, CEM 2015 (pp. 4-5). Article 7237412 (2015 Computational Electromagnetics International Workshop, CEM 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/CEM.2015.7237412

Transitioning HPC software to exascale heterogeneous computing. / Hwu, Wen Mei; Chang, Li Wen; Kim, Hee Seok et al.
2015 Computational Electromagnetics International Workshop, CEM 2015. Institute of Electrical and Electronics Engineers Inc., 2015. p. 4-5 7237412 (2015 Computational Electromagnetics International Workshop, CEM 2015).

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

Hwu, WM, Chang, LW, Kim, HS, Dakkak, A & El Hajj, I 2015, Transitioning HPC software to exascale heterogeneous computing. in 2015 Computational Electromagnetics International Workshop, CEM 2015., 7237412, 2015 Computational Electromagnetics International Workshop, CEM 2015, Institute of Electrical and Electronics Engineers Inc., pp. 4-5, Computational Electromagnetics International Workshop, CEM 2015, Izmir, Turkey, 7/1/15. https://doi.org/10.1109/CEM.2015.7237412

@inproceedings{243754c6f49d4bb7bd9e95bef815d03b,

title = "Transitioning HPC software to exascale heterogeneous computing",

abstract = "An increasing portion of the top supercomputers in the world, including Blue Waters, have heterogeneous CPUGPU computational units. As we move towards exascale, we can expect even more pervasive deployment of heterogeneous computational units. While a handful of science teams can already use heterogeneous computational units in their production applications, there is still significant room for the growing use. This paper presents the current state and projected path for transitioning software into this new paradigm. We first summarize the currently practical languages such as OpenCL, OpenACC, and C++AMP, in increasing levels of productivity, highlighting their recent advancements in supporting performance portability and maintainability. We will then give a brief overview of some emerging programming systems such as TANGRAM and Troilet that are designed to further enhance developer productivity for heterogeneous computing.",

author = "Hwu, {Wen Mei} and Chang, {Li Wen} and Kim, {Hee Seok} and Abdul Dakkak and {El Hajj}, Izzat",

note = "Publisher Copyright: {\textcopyright} 2015 IEEE.; Computational Electromagnetics International Workshop, CEM 2015 ; Conference date: 01-07-2015 Through 04-07-2015",

year = "2015",

month = sep,

day = "2",

doi = "10.1109/CEM.2015.7237412",

language = "English (US)",

series = "2015 Computational Electromagnetics International Workshop, CEM 2015",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

pages = "4--5",

booktitle = "2015 Computational Electromagnetics International Workshop, CEM 2015",

address = "United States",

}

TY - GEN

T1 - Transitioning HPC software to exascale heterogeneous computing

AU - Hwu, Wen Mei

AU - Chang, Li Wen

AU - Kim, Hee Seok

AU - Dakkak, Abdul

AU - El Hajj, Izzat

PY - 2015/9/2

Y1 - 2015/9/2

N2 - An increasing portion of the top supercomputers in the world, including Blue Waters, have heterogeneous CPUGPU computational units. As we move towards exascale, we can expect even more pervasive deployment of heterogeneous computational units. While a handful of science teams can already use heterogeneous computational units in their production applications, there is still significant room for the growing use. This paper presents the current state and projected path for transitioning software into this new paradigm. We first summarize the currently practical languages such as OpenCL, OpenACC, and C++AMP, in increasing levels of productivity, highlighting their recent advancements in supporting performance portability and maintainability. We will then give a brief overview of some emerging programming systems such as TANGRAM and Troilet that are designed to further enhance developer productivity for heterogeneous computing.

AB - An increasing portion of the top supercomputers in the world, including Blue Waters, have heterogeneous CPUGPU computational units. As we move towards exascale, we can expect even more pervasive deployment of heterogeneous computational units. While a handful of science teams can already use heterogeneous computational units in their production applications, there is still significant room for the growing use. This paper presents the current state and projected path for transitioning software into this new paradigm. We first summarize the currently practical languages such as OpenCL, OpenACC, and C++AMP, in increasing levels of productivity, highlighting their recent advancements in supporting performance portability and maintainability. We will then give a brief overview of some emerging programming systems such as TANGRAM and Troilet that are designed to further enhance developer productivity for heterogeneous computing.

UR - http://www.scopus.com/inward/record.url?scp=84957900340&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84957900340&partnerID=8YFLogxK

U2 - 10.1109/CEM.2015.7237412

DO - 10.1109/CEM.2015.7237412

M3 - Conference contribution

AN - SCOPUS:84957900340

T3 - 2015 Computational Electromagnetics International Workshop, CEM 2015

SP - 4

EP - 5

BT - 2015 Computational Electromagnetics International Workshop, CEM 2015

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - Computational Electromagnetics International Workshop, CEM 2015

Y2 - 1 July 2015 through 4 July 2015

ER -

Transitioning HPC software to exascale heterogeneous computing

Abstract

Publication series

Other

ASJC Scopus subject areas

Online availability

Library availability

Related links

Fingerprint

Cite this