pupyMPI - MPI implemented in pure Python

Rune Bromer; Frederik Hantho; Brian Vinter

doi:10.1007/978-3-642-24449-0_16

pupyMPI - MPI implemented in pure Python

Rune Bromer^*, Frederik Hantho, Brian Vinter

^*Corresponding author af dette arbejde

1 Citationer (Scopus)

Abstract

As distributed memory systems have become common, the de facto standard for communication is still the Message Passing Interface (MPI). pupyMPI is a pure Python implementation of a broad subset of the MPI 1.3 specifications that allows Python programmers to utilize multiple CPUs with datatypes and memory handled transparently. pupyMPI also implements a few non-standard extensions such as non-blocking collectives and the option of suspending, migrating and resuming the distributed computation of a pupyMPI program. This paper introduces pupyMPI and presents benchmarks against C implementations of MPI, which show acceptable performance.

Originalsprog	Engelsk
Titel	Recent Advances in the Message Passing Interface : 18th European MPI Users’ Group Meeting, EuroMPI 2011, Santorini, Greece, September 18-21, 2011. Proceedings
Antal sider	10
Forlag	Springer
Publikationsdato	2011
Sider	130-139
ISBN (Trykt)	978-3-642-24448-3
ISBN (Elektronisk)	978-3-642-24449-0
DOI	https://doi.org/10.1007/978-3-642-24449-0_16
Status	Udgivet - 2011
Begivenhed	18th European MPI Users' Group Meeting: recent advances in the message passing interface - Santorini, Grækenland Varighed: 18 sep. 2011 → 21 sep. 2011 Konferencens nummer: 18

Konference

Konference	18th European MPI Users' Group Meeting
Nummer	18
Land/Område	Grækenland
By	Santorini
Periode	18/09/2011 → 21/09/2011

Navn	Lecture notes in computer science
Vol/bind	6960
ISSN	0302-9743

Adgang til dokumentet

10.1007/978-3-642-24449-0_16

Andre filer og links

Link to publication in Scopus

Citationsformater

pupyMPI - MPI implemented in pure Python. / Bromer, Rune; Hantho, Frederik; Vinter, Brian.

Recent Advances in the Message Passing Interface: 18th European MPI Users’ Group Meeting, EuroMPI 2011, Santorini, Greece, September 18-21, 2011. Proceedings. Springer, 2011. s. 130-139 (Lecture notes in computer science, Bind 6960).

Publikation: Bidrag til bog/antologi/rapport › Konferencebidrag i proceedings › Forskning › peer review

Bromer, R, Hantho, F & Vinter, B 2011, pupyMPI - MPI implemented in pure Python. i Recent Advances in the Message Passing Interface: 18th European MPI Users’ Group Meeting, EuroMPI 2011, Santorini, Greece, September 18-21, 2011. Proceedings. Springer, Lecture notes in computer science, bind 6960, s. 130-139, 18th European MPI Users' Group Meeting, Santorini, Grækenland, 18/09/2011. https://doi.org/10.1007/978-3-642-24449-0_16

@inproceedings{b9071d3477774ab6a3addb81fa5af11d,

title = "pupyMPI - MPI implemented in pure Python",

abstract = "As distributed memory systems have become common, the de facto standard for communication is still the Message Passing Interface (MPI). pupyMPI is a pure Python implementation of a broad subset of the MPI 1.3 specifications that allows Python programmers to utilize multiple CPUs with datatypes and memory handled transparently. pupyMPI also implements a few non-standard extensions such as non-blocking collectives and the option of suspending, migrating and resuming the distributed computation of a pupyMPI program. This paper introduces pupyMPI and presents benchmarks against C implementations of MPI, which show acceptable performance.",

keywords = "high-level languages, MPI, parallel computing, Python, scientific computing",

author = "Rune Bromer and Frederik Hantho and Brian Vinter",

year = "2011",

doi = "10.1007/978-3-642-24449-0_16",

language = "English",

isbn = "978-3-642-24448-3",

series = "Lecture notes in computer science",

publisher = "Springer",

pages = "130--139",

booktitle = "Recent Advances in the Message Passing Interface",

note = "18th European MPI Users' Group Meeting ; Conference date: 18-09-2011 Through 21-09-2011",

}

TY - GEN

T1 - pupyMPI - MPI implemented in pure Python

AU - Bromer, Rune

AU - Hantho, Frederik

AU - Vinter, Brian

N1 - Conference code: 18

PY - 2011

Y1 - 2011

N2 - As distributed memory systems have become common, the de facto standard for communication is still the Message Passing Interface (MPI). pupyMPI is a pure Python implementation of a broad subset of the MPI 1.3 specifications that allows Python programmers to utilize multiple CPUs with datatypes and memory handled transparently. pupyMPI also implements a few non-standard extensions such as non-blocking collectives and the option of suspending, migrating and resuming the distributed computation of a pupyMPI program. This paper introduces pupyMPI and presents benchmarks against C implementations of MPI, which show acceptable performance.

AB - As distributed memory systems have become common, the de facto standard for communication is still the Message Passing Interface (MPI). pupyMPI is a pure Python implementation of a broad subset of the MPI 1.3 specifications that allows Python programmers to utilize multiple CPUs with datatypes and memory handled transparently. pupyMPI also implements a few non-standard extensions such as non-blocking collectives and the option of suspending, migrating and resuming the distributed computation of a pupyMPI program. This paper introduces pupyMPI and presents benchmarks against C implementations of MPI, which show acceptable performance.

KW - high-level languages

KW - MPI

KW - parallel computing

KW - Python

KW - scientific computing

UR - http://www.scopus.com/inward/record.url?scp=80053008522&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-24449-0_16

DO - 10.1007/978-3-642-24449-0_16

M3 - Article in proceedings

AN - SCOPUS:80053008522

SN - 978-3-642-24448-3

T3 - Lecture notes in computer science

SP - 130

EP - 139

BT - Recent Advances in the Message Passing Interface

PB - Springer

T2 - 18th European MPI Users' Group Meeting

Y2 - 18 September 2011 through 21 September 2011

ER -

pupyMPI - MPI implemented in pure Python

Abstract

Konference

Adgang til dokumentet

Andre filer og links

Fingeraftryk

Citationsformater