Using hybrid MPI and OpenMP programming to optimize communications in parallel loop self-scheduling schemes for multicore PC clusters

Chao-Chin Wu, Lien-Fu Lai, Chao Tung Yang, Po Hsun Chiu

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Recently, a series of parallel loop self-scheduling schemes have been proposed, especially for heterogeneous cluster systems. However, they employed the MPI programming model to construct the applications without considering whether the computing node is multicore architecture or not. As a result, every processor core has to communicate directly with the master node for requesting new tasks no matter the fact that the processor cores on the same node can communicate with each other through the underlying shared memory. To address the problem of higher communication overhead, in this paper we propose to adopt hybrid MPI and OpenMP programming model to design two-level parallel loop self-scheduling schemes. In the first level, each computing node runs an MPI process for inter-node communications. In the second level, each processor core runs an OpenMP thread to execute the iterations assigned for its resident node. Experimental results show that our method outperforms the previous works.

Original languageEnglish
Pages (from-to)31-61
Number of pages31
JournalJournal of Supercomputing
Volume60
Issue number1
DOIs
Publication statusPublished - 2012 Apr 1

Fingerprint

PC Cluster
OpenMP
Programming
Scheduling
Optimise
Communication
Vertex of a graph
Data storage equipment
Programming Model
Computing
Shared Memory
Thread
Iteration
Series
Experimental Results

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Software
  • Information Systems
  • Hardware and Architecture

Cite this

@article{01ba60b642c146ca9cebbfcf29a2b0f5,
title = "Using hybrid MPI and OpenMP programming to optimize communications in parallel loop self-scheduling schemes for multicore PC clusters",
abstract = "Recently, a series of parallel loop self-scheduling schemes have been proposed, especially for heterogeneous cluster systems. However, they employed the MPI programming model to construct the applications without considering whether the computing node is multicore architecture or not. As a result, every processor core has to communicate directly with the master node for requesting new tasks no matter the fact that the processor cores on the same node can communicate with each other through the underlying shared memory. To address the problem of higher communication overhead, in this paper we propose to adopt hybrid MPI and OpenMP programming model to design two-level parallel loop self-scheduling schemes. In the first level, each computing node runs an MPI process for inter-node communications. In the second level, each processor core runs an OpenMP thread to execute the iterations assigned for its resident node. Experimental results show that our method outperforms the previous works.",
author = "Chao-Chin Wu and Lien-Fu Lai and Yang, {Chao Tung} and Chiu, {Po Hsun}",
year = "2012",
month = "4",
day = "1",
doi = "10.1007/s11227-009-0271-z",
language = "English",
volume = "60",
pages = "31--61",
journal = "Journal of Supercomputing",
issn = "0920-8542",
publisher = "Springer Netherlands",
number = "1",

}

Using hybrid MPI and OpenMP programming to optimize communications in parallel loop self-scheduling schemes for multicore PC clusters. / Wu, Chao-Chin; Lai, Lien-Fu; Yang, Chao Tung; Chiu, Po Hsun.

In: Journal of Supercomputing, Vol. 60, No. 1, 01.04.2012, p. 31-61.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Using hybrid MPI and OpenMP programming to optimize communications in parallel loop self-scheduling schemes for multicore PC clusters

AU - Wu, Chao-Chin

AU - Lai, Lien-Fu

AU - Yang, Chao Tung

AU - Chiu, Po Hsun

PY - 2012/4/1

Y1 - 2012/4/1

N2 - Recently, a series of parallel loop self-scheduling schemes have been proposed, especially for heterogeneous cluster systems. However, they employed the MPI programming model to construct the applications without considering whether the computing node is multicore architecture or not. As a result, every processor core has to communicate directly with the master node for requesting new tasks no matter the fact that the processor cores on the same node can communicate with each other through the underlying shared memory. To address the problem of higher communication overhead, in this paper we propose to adopt hybrid MPI and OpenMP programming model to design two-level parallel loop self-scheduling schemes. In the first level, each computing node runs an MPI process for inter-node communications. In the second level, each processor core runs an OpenMP thread to execute the iterations assigned for its resident node. Experimental results show that our method outperforms the previous works.

AB - Recently, a series of parallel loop self-scheduling schemes have been proposed, especially for heterogeneous cluster systems. However, they employed the MPI programming model to construct the applications without considering whether the computing node is multicore architecture or not. As a result, every processor core has to communicate directly with the master node for requesting new tasks no matter the fact that the processor cores on the same node can communicate with each other through the underlying shared memory. To address the problem of higher communication overhead, in this paper we propose to adopt hybrid MPI and OpenMP programming model to design two-level parallel loop self-scheduling schemes. In the first level, each computing node runs an MPI process for inter-node communications. In the second level, each processor core runs an OpenMP thread to execute the iterations assigned for its resident node. Experimental results show that our method outperforms the previous works.

UR - http://www.scopus.com/inward/record.url?scp=84858079019&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84858079019&partnerID=8YFLogxK

U2 - 10.1007/s11227-009-0271-z

DO - 10.1007/s11227-009-0271-z

M3 - Article

AN - SCOPUS:84858079019

VL - 60

SP - 31

EP - 61

JO - Journal of Supercomputing

JF - Journal of Supercomputing

SN - 0920-8542

IS - 1

ER -