Parsing in parallel on multiple cores and GPUs

Mark Johnson

Parsing in parallel on multiple cores and GPUs

Mark Johnson

School of Computing

Research output: Contribution to journal › Conference paper › peer-review

12 Citations (Scopus)

118 Downloads (Pure)

Abstract

This paper examines the ways in which parallelism can be used to speed the parsing of dense PCFGs. We focus on two kinds of parallelism here: Symmetric Multi-Processing (SMP) parallelism on shared-memory multicore CPUs, and Single-Instruction Multiple- Thread (SIMT) parallelism on GPUs. We describe how to achieve speed-ups over an already very efficient baseline parser using both kinds of technology. For our dense PCFG parsing task we obtained a 60×speed-up using SMP and SSE parallelism coupled with a cache-sensitive algorithm design, parsing section 24 of the Penn WSJ treebank in a little over 2 secs.

Original language	English
Pages (from-to)	29-37
Number of pages	9
Journal	Proceedings of the Australasian Language Technology Association Workshop 2011
Publication status	Published - 2011
Event	Australasian Language Technology Workshop (9th : 2011) - Canberra Duration: 1 Dec 2011 → 2 Dec 2011

Bibliographical note

Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

Access to Document

Publisher Version (open access)Final published version, 6.66 MBLicence: Unspecified

Cite this

@article{bc8ebe39cff8460582145f8436be1daa,

title = "Parsing in parallel on multiple cores and GPUs",

abstract = "This paper examines the ways in which parallelism can be used to speed the parsing of dense PCFGs. We focus on two kinds of parallelism here: Symmetric Multi-Processing (SMP) parallelism on shared-memory multicore CPUs, and Single-Instruction Multiple- Thread (SIMT) parallelism on GPUs. We describe how to achieve speed-ups over an already very efficient baseline parser using both kinds of technology. For our dense PCFG parsing task we obtained a 60×speed-up using SMP and SSE parallelism coupled with a cache-sensitive algorithm design, parsing section 24 of the Penn WSJ treebank in a little over 2 secs.",

author = "Mark Johnson",

note = "Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.; Australasian Language Technology Workshop (9th : 2011) ; Conference date: 01-12-2011 Through 02-12-2011",

year = "2011",

language = "English",

pages = "29--37",

journal = "Proceedings of the Australasian Language Technology Association Workshop 2011",

issn = "1834-7037",

publisher = "Australian Language Technology Association",

}

TY - JOUR

T1 - Parsing in parallel on multiple cores and GPUs

AU - Johnson, Mark

N1 - Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.

PY - 2011

Y1 - 2011

N2 - This paper examines the ways in which parallelism can be used to speed the parsing of dense PCFGs. We focus on two kinds of parallelism here: Symmetric Multi-Processing (SMP) parallelism on shared-memory multicore CPUs, and Single-Instruction Multiple- Thread (SIMT) parallelism on GPUs. We describe how to achieve speed-ups over an already very efficient baseline parser using both kinds of technology. For our dense PCFG parsing task we obtained a 60×speed-up using SMP and SSE parallelism coupled with a cache-sensitive algorithm design, parsing section 24 of the Penn WSJ treebank in a little over 2 secs.

AB - This paper examines the ways in which parallelism can be used to speed the parsing of dense PCFGs. We focus on two kinds of parallelism here: Symmetric Multi-Processing (SMP) parallelism on shared-memory multicore CPUs, and Single-Instruction Multiple- Thread (SIMT) parallelism on GPUs. We describe how to achieve speed-ups over an already very efficient baseline parser using both kinds of technology. For our dense PCFG parsing task we obtained a 60×speed-up using SMP and SSE parallelism coupled with a cache-sensitive algorithm design, parsing section 24 of the Penn WSJ treebank in a little over 2 secs.

M3 - Conference paper

SN - 1834-7037

SP - 29

EP - 37

JO - Proceedings of the Australasian Language Technology Association Workshop 2011

JF - Proceedings of the Australasian Language Technology Association Workshop 2011

T2 - Australasian Language Technology Workshop (9th : 2011)

Y2 - 1 December 2011 through 2 December 2011

ER -

Parsing in parallel on multiple cores and GPUs

Abstract

Bibliographical note

Access to Document

Fingerprint

Cite this