A Case Study of OpenMP applied to Map/Reduce-style Computations

Mahwish Arif; Hans Vandierendonck

doi:10.1007/978-3-319-24595-9_12

A Case Study of OpenMP applied to Map/Reduce-style Computations

Mahwish Arif, Hans Vandierendonck

School of Electronics, Electrical Engineering and Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Chapter (peer-reviewed) › peer-review

3 Citations (Scopus)

2130 Downloads (Pure)

Abstract

As data analytics are growing in importance they are also quickly becoming one of the dominant application domains that require parallel processing. This paper investigates the applicability of OpenMP, the dominant shared-memory parallel programming model in high-performance computing, to the domain of data analytics. We contrast the performance and programmability of key data analytics benchmarks against Phoenix++, a state-of-the-art shared memory map/reduce programming system. Our study shows that OpenMP outperforms the Phoenix++ system by a large margin for several benchmarks. In other cases, however, the programming model is lacking support for this application domain.

Original language	English
Title of host publication	OpenMP: Heterogenous Execution and Data Movements
Pages	162-174
Volume	9342
ISBN (Electronic)	978-3-319-24595-9
DOIs	https://doi.org/10.1007/978-3-319-24595-9_12
Publication status	Published - Nov 2015
Event	11th International Workshop on OpenMP, IWOMP 2015 - Aachen, Germany Duration: 01 Oct 2015 → 02 Oct 2015 http://www.iwomp.org/

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer
ISSN (Print)	0302-9743

Conference

Conference	11th International Workshop on OpenMP, IWOMP 2015
Country/Territory	Germany
City	Aachen
Period	01/10/2015 → 02/10/2015
Internet address	http://www.iwomp.org/

Keywords

OpenMP
map/reduce
REDUCTION

Access to Document

10.1007/978-3-319-24595-9_12Licence: Unspecified

A case study of OpenMP Applied to MapReduce-style Computations
Copyright Springer 2015. The final publication is available at Springer via http://dx.doi.org/10.1007/978-3-319-24595-9_12.
Accepted author manuscript, 1.12 MBLicence: Unspecified

2 Finished

R1451CSC: Hybrid Static/Dynamic Scheduling for Task Dataflow Parallel Programs
Vandierendonck, H.
28/07/2014 → 02/03/2017
Project: Research
R6438CSC: An Adaptive, highly Scalable Analytics Platform
Vandierendonck, H., Nikolopoulos, D. & Robinson, P.
21/03/2014 → 28/02/2017
Project: Research

Hans Vandierendonck
- h.vandierendonckqub.acuk
Person: Academic

Cite this

@inbook{317403e603ab443eb429907160e30867,

title = "A Case Study of OpenMP applied to Map/Reduce-style Computations",

abstract = "As data analytics are growing in importance they are also quickly becoming one of the dominant application domains that require parallel processing. This paper investigates the applicability of OpenMP, the dominant shared-memory parallel programming model in high-performance computing, to the domain of data analytics. We contrast the performance and programmability of key data analytics benchmarks against Phoenix++, a state-of-the-art shared memory map/reduce programming system. Our study shows that OpenMP outperforms the Phoenix++ system by a large margin for several benchmarks. In other cases, however, the programming model is lacking support for this application domain.",

keywords = "OpenMP, map/reduce, REDUCTION",

author = "Mahwish Arif and Hans Vandierendonck",

year = "2015",

month = nov,

doi = "10.1007/978-3-319-24595-9_12",

language = "English",

isbn = "978-3-319-24594-2",

volume = "9342",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "162--174",

booktitle = "OpenMP: Heterogenous Execution and Data Movements",

note = "11th International Workshop on OpenMP, IWOMP 2015 ; Conference date: 01-10-2015 Through 02-10-2015",

url = "http://www.iwomp.org/",

}

TY - CHAP

T1 - A Case Study of OpenMP applied to Map/Reduce-style Computations

AU - Arif, Mahwish

AU - Vandierendonck, Hans

PY - 2015/11

Y1 - 2015/11

N2 - As data analytics are growing in importance they are also quickly becoming one of the dominant application domains that require parallel processing. This paper investigates the applicability of OpenMP, the dominant shared-memory parallel programming model in high-performance computing, to the domain of data analytics. We contrast the performance and programmability of key data analytics benchmarks against Phoenix++, a state-of-the-art shared memory map/reduce programming system. Our study shows that OpenMP outperforms the Phoenix++ system by a large margin for several benchmarks. In other cases, however, the programming model is lacking support for this application domain.

AB - As data analytics are growing in importance they are also quickly becoming one of the dominant application domains that require parallel processing. This paper investigates the applicability of OpenMP, the dominant shared-memory parallel programming model in high-performance computing, to the domain of data analytics. We contrast the performance and programmability of key data analytics benchmarks against Phoenix++, a state-of-the-art shared memory map/reduce programming system. Our study shows that OpenMP outperforms the Phoenix++ system by a large margin for several benchmarks. In other cases, however, the programming model is lacking support for this application domain.

KW - OpenMP

KW - map/reduce

KW - REDUCTION

U2 - 10.1007/978-3-319-24595-9_12

DO - 10.1007/978-3-319-24595-9_12

M3 - Chapter (peer-reviewed)

SN - 978-3-319-24594-2

VL - 9342

T3 - Lecture Notes in Computer Science

SP - 162

EP - 174

BT - OpenMP: Heterogenous Execution and Data Movements

T2 - 11th International Workshop on OpenMP, IWOMP 2015

Y2 - 1 October 2015 through 2 October 2015

ER -

A Case Study of OpenMP applied to Map/Reduce-style Computations

Abstract

Publication series

Conference

Keywords

Access to Document

Fingerprint

Projects

R1451CSC: Hybrid Static/Dynamic Scheduling for Task Dataflow Parallel Programs

R6438CSC: An Adaptive, highly Scalable Analytics Platform

Profiles

Hans Vandierendonck

Cite this