An efficient collective communication method for Grid scale networks

Kyung Lang Park, Hwang Jik Lee, Youn Joo Lee, Oh Young Kwon, Sung Yong Park, Hyung Woo Park, Shin-Dug Kim

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

This research is to design a collection of effective collective operations for the Grid scale network environment. In the Grid, several network features should be specified and adapted into the algorithmic design of collective operations. For this, we characterize the Grid and design hierarchical latency optimal tree algorithm for the MPI (message passing interface) library. The experimental results show that performance gain in performing collective operations can be achieved by around 160% compared with MPICH-G2 which is based on the flat algorithm.

Original languageEnglish
Pages (from-to)819-828
Number of pages10
JournalLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume2660
Publication statusPublished - 2003 Dec 1

Fingerprint

Collective Communication
Grid
Communication
Message Passing Interface
Tree Algorithms
Message passing
Optimal Algorithm
Latency
Experimental Results
Design

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Park, Kyung Lang ; Lee, Hwang Jik ; Lee, Youn Joo ; Kwon, Oh Young ; Park, Sung Yong ; Park, Hyung Woo ; Kim, Shin-Dug. / An efficient collective communication method for Grid scale networks. In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 2003 ; Vol. 2660. pp. 819-828.
@article{4e907ad145d44543b0b4f07405f16c3f,
title = "An efficient collective communication method for Grid scale networks",
abstract = "This research is to design a collection of effective collective operations for the Grid scale network environment. In the Grid, several network features should be specified and adapted into the algorithmic design of collective operations. For this, we characterize the Grid and design hierarchical latency optimal tree algorithm for the MPI (message passing interface) library. The experimental results show that performance gain in performing collective operations can be achieved by around 160{\%} compared with MPICH-G2 which is based on the flat algorithm.",
author = "Park, {Kyung Lang} and Lee, {Hwang Jik} and Lee, {Youn Joo} and Kwon, {Oh Young} and Park, {Sung Yong} and Park, {Hyung Woo} and Shin-Dug Kim",
year = "2003",
month = "12",
day = "1",
language = "English",
volume = "2660",
pages = "819--828",
journal = "Lecture Notes in Computer Science",
issn = "0302-9743",
publisher = "Springer Verlag",

}

An efficient collective communication method for Grid scale networks. / Park, Kyung Lang; Lee, Hwang Jik; Lee, Youn Joo; Kwon, Oh Young; Park, Sung Yong; Park, Hyung Woo; Kim, Shin-Dug.

In: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Vol. 2660, 01.12.2003, p. 819-828.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An efficient collective communication method for Grid scale networks

AU - Park, Kyung Lang

AU - Lee, Hwang Jik

AU - Lee, Youn Joo

AU - Kwon, Oh Young

AU - Park, Sung Yong

AU - Park, Hyung Woo

AU - Kim, Shin-Dug

PY - 2003/12/1

Y1 - 2003/12/1

N2 - This research is to design a collection of effective collective operations for the Grid scale network environment. In the Grid, several network features should be specified and adapted into the algorithmic design of collective operations. For this, we characterize the Grid and design hierarchical latency optimal tree algorithm for the MPI (message passing interface) library. The experimental results show that performance gain in performing collective operations can be achieved by around 160% compared with MPICH-G2 which is based on the flat algorithm.

AB - This research is to design a collection of effective collective operations for the Grid scale network environment. In the Grid, several network features should be specified and adapted into the algorithmic design of collective operations. For this, we characterize the Grid and design hierarchical latency optimal tree algorithm for the MPI (message passing interface) library. The experimental results show that performance gain in performing collective operations can be achieved by around 160% compared with MPICH-G2 which is based on the flat algorithm.

UR - http://www.scopus.com/inward/record.url?scp=35248825533&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=35248825533&partnerID=8YFLogxK

M3 - Article

VL - 2660

SP - 819

EP - 828

JO - Lecture Notes in Computer Science

JF - Lecture Notes in Computer Science

SN - 0302-9743

ER -