The Loop-of-Stencil-Reduce Paradigm

Marco Aldinucci; Marco Danelutto; Maurizio Drocco; Peter Kilpatrick; Guilherme Peretti Pezzi; Massimo Torquati

doi:10.1109/Trustcom.2015.628

The Loop-of-Stencil-Reduce Paradigm

Marco Aldinucci, Marco Danelutto, Maurizio Drocco, Peter Kilpatrick, Guilherme Peretti Pezzi, Massimo Torquati

School of Electronics, Electrical Engineering and Computer Science

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

7 Citations (Scopus)

Abstract

In this paper we advocate the Loop-of-stencil-reduce pattern as a way to simplify the parallel programming of heterogeneous platforms (multicore+GPUs). Loop-of-Stencil-reduce is general enough to subsume map, reduce, map-reduce, stencil, stencil-reduce, and, crucially, their usage in a loop. It transparently targets (by using OpenCL) combinations of CPU cores and GPUs, and it makes it possible to simplify the deployment of a single stencil computation kernel on different GPUs. The paper discusses the implementation of Loop-of-stencil-reduce within the FastFlow parallel framework, considering a simple iterative data-parallel application as running example (Game of Life) and a highly effective parallel filter for visual data restoration to assess performance. Thanks to the high-level design of the Loop-of-stencil-reduce, it was possible to run the filter seamlessly on a multicore machine, on multi-GPUs, and on both.

Original language	English
Title of host publication	Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	172-177
Number of pages	6
Volume	3
ISBN (Print)	9781467379519
DOIs	https://doi.org/10.1109/Trustcom.2015.628
Publication status	Published - 02 Dec 2015
Event	14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015 - Helsinki, Finland Duration: 20 Aug 2015 → 22 Aug 2015

Conference

Conference	14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015
Country/Territory	Finland
City	Helsinki
Period	20/08/2015 → 22/08/2015

Keywords

fastflow
GPUs
heterogeneous platforms
multi-core
OpenCL
parallel patterns
skeletons

ASJC Scopus subject areas

Computer Networks and Communications

Access to Document

10.1109/Trustcom.2015.628

Cite this

Aldinucci, M., Danelutto, M., Drocco, M., Kilpatrick, P., Pezzi, G. P., & Torquati, M. (2015). The Loop-of-Stencil-Reduce Paradigm. In Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015 (Vol. 3, pp. 172-177). Article 7345644 Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/Trustcom.2015.628

@inproceedings{b54390895d5c412798f3c983a2c14de4,

title = "The Loop-of-Stencil-Reduce Paradigm",

abstract = "In this paper we advocate the Loop-of-stencil-reduce pattern as a way to simplify the parallel programming of heterogeneous platforms (multicore+GPUs). Loop-of-Stencil-reduce is general enough to subsume map, reduce, map-reduce, stencil, stencil-reduce, and, crucially, their usage in a loop. It transparently targets (by using OpenCL) combinations of CPU cores and GPUs, and it makes it possible to simplify the deployment of a single stencil computation kernel on different GPUs. The paper discusses the implementation of Loop-of-stencil-reduce within the FastFlow parallel framework, considering a simple iterative data-parallel application as running example (Game of Life) and a highly effective parallel filter for visual data restoration to assess performance. Thanks to the high-level design of the Loop-of-stencil-reduce, it was possible to run the filter seamlessly on a multicore machine, on multi-GPUs, and on both.",

keywords = "fastflow, GPUs, heterogeneous platforms, multi-core, OpenCL, parallel patterns, skeletons",

author = "Marco Aldinucci and Marco Danelutto and Maurizio Drocco and Peter Kilpatrick and Pezzi, {Guilherme Peretti} and Massimo Torquati",

year = "2015",

month = dec,

day = "2",

doi = "10.1109/Trustcom.2015.628",

language = "English",

isbn = "9781467379519",

volume = "3",

pages = "172--177",

booktitle = "Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015",

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

address = "United States",

note = "14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015 ; Conference date: 20-08-2015 Through 22-08-2015",

}

Aldinucci, M, Danelutto, M, Drocco, M, Kilpatrick, P, Pezzi, GP & Torquati, M 2015, The Loop-of-Stencil-Reduce Paradigm. in Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015. vol. 3, 7345644, Institute of Electrical and Electronics Engineers Inc., pp. 172-177, 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015, Helsinki, Finland, 20/08/2015. https://doi.org/10.1109/Trustcom.2015.628

The Loop-of-Stencil-Reduce Paradigm. / Aldinucci, Marco; Danelutto, Marco; Drocco, Maurizio et al.
Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015. Vol. 3 Institute of Electrical and Electronics Engineers Inc., 2015. p. 172-177 7345644.

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

TY - GEN

T1 - The Loop-of-Stencil-Reduce Paradigm

AU - Aldinucci, Marco

AU - Danelutto, Marco

AU - Drocco, Maurizio

AU - Kilpatrick, Peter

AU - Pezzi, Guilherme Peretti

AU - Torquati, Massimo

PY - 2015/12/2

Y1 - 2015/12/2

N2 - In this paper we advocate the Loop-of-stencil-reduce pattern as a way to simplify the parallel programming of heterogeneous platforms (multicore+GPUs). Loop-of-Stencil-reduce is general enough to subsume map, reduce, map-reduce, stencil, stencil-reduce, and, crucially, their usage in a loop. It transparently targets (by using OpenCL) combinations of CPU cores and GPUs, and it makes it possible to simplify the deployment of a single stencil computation kernel on different GPUs. The paper discusses the implementation of Loop-of-stencil-reduce within the FastFlow parallel framework, considering a simple iterative data-parallel application as running example (Game of Life) and a highly effective parallel filter for visual data restoration to assess performance. Thanks to the high-level design of the Loop-of-stencil-reduce, it was possible to run the filter seamlessly on a multicore machine, on multi-GPUs, and on both.

AB - In this paper we advocate the Loop-of-stencil-reduce pattern as a way to simplify the parallel programming of heterogeneous platforms (multicore+GPUs). Loop-of-Stencil-reduce is general enough to subsume map, reduce, map-reduce, stencil, stencil-reduce, and, crucially, their usage in a loop. It transparently targets (by using OpenCL) combinations of CPU cores and GPUs, and it makes it possible to simplify the deployment of a single stencil computation kernel on different GPUs. The paper discusses the implementation of Loop-of-stencil-reduce within the FastFlow parallel framework, considering a simple iterative data-parallel application as running example (Game of Life) and a highly effective parallel filter for visual data restoration to assess performance. Thanks to the high-level design of the Loop-of-stencil-reduce, it was possible to run the filter seamlessly on a multicore machine, on multi-GPUs, and on both.

KW - fastflow

KW - GPUs

KW - heterogeneous platforms

KW - multi-core

KW - OpenCL

KW - parallel patterns

KW - skeletons

UR - http://www.scopus.com/inward/record.url?scp=84969262578&partnerID=8YFLogxK

U2 - 10.1109/Trustcom.2015.628

DO - 10.1109/Trustcom.2015.628

M3 - Conference contribution

AN - SCOPUS:84969262578

SN - 9781467379519

VL - 3

SP - 172

EP - 177

BT - Proceedings - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 14th IEEE International Conference on Trust, Security and Privacy in Computing and Communications, TrustCom 2015

Y2 - 20 August 2015 through 22 August 2015

ER -

The Loop-of-Stencil-Reduce Paradigm

Abstract

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this