Performance evaluation of enhancement of the layered self-scheduling approach for heterogeneous multicore cluster systems

Chao-Chin Wu, Lien-Fu Lai, Liang Tsung Huang, Minglung Chen

Research output: Contribution to journalArticle

Abstract

Previously we have proposed a Layered Self-Scheduling (LSS) approach that is a hybrid MPI and OpenMP based loop self-scheduling approach for dealing with the heterogeneity problem on a cluster system consisting of multi-core compute nodes, where the allocation functions of several well-known schemes have been modified for better performance. Though LSS provides better performance than the conventional self-scheduling schemes, we found the performance can be improved further after our comprehensive experiments and analyses. The newly proposed task scheduling strategy, called Enhanced Layered Self-Scheduling (ELSS), aims at how to utilize the compute powers of multiple processor cores more efficiently in the master compute node and how to schedule tasks to have more stable performance improvements. We have evaluated the new task scheduling strategy by three benchmark applications: Matrix Multiplication, Monte Carlo Integration, and Mandelbrot Set Computation. It is recommended that the global scheduler adopts Guided Self-Scheduling (GSS) for all, and the local scheduler adopts the static scheme for applications with regular workload distribution but any scheme for applications with irregular workload distribution. Experimental results show the best speedups obtained by ELSS for the three benchmark programs are 1.373, 13.34 and 2.4, respectively, compared with that scheduled by LSS.

Original languageEnglish
Pages (from-to)399-430
Number of pages32
JournalJournal of Supercomputing
Volume62
Issue number1
DOIs
Publication statusPublished - 2012 Oct 1

Fingerprint

Performance Evaluation
Enhancement
Scheduling
Task Scheduling
Scheduler
Workload
Benchmark
Mandelbrot set
Monte Carlo Integration
OpenMP
Matrix multiplication
Vertex of a graph
Irregular
Schedule
Experimental Results
Experiment

All Science Journal Classification (ASJC) codes

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

Cite this

@article{8d89d9a383c649d0a9d55fdeb87f5f0c,
title = "Performance evaluation of enhancement of the layered self-scheduling approach for heterogeneous multicore cluster systems",
abstract = "Previously we have proposed a Layered Self-Scheduling (LSS) approach that is a hybrid MPI and OpenMP based loop self-scheduling approach for dealing with the heterogeneity problem on a cluster system consisting of multi-core compute nodes, where the allocation functions of several well-known schemes have been modified for better performance. Though LSS provides better performance than the conventional self-scheduling schemes, we found the performance can be improved further after our comprehensive experiments and analyses. The newly proposed task scheduling strategy, called Enhanced Layered Self-Scheduling (ELSS), aims at how to utilize the compute powers of multiple processor cores more efficiently in the master compute node and how to schedule tasks to have more stable performance improvements. We have evaluated the new task scheduling strategy by three benchmark applications: Matrix Multiplication, Monte Carlo Integration, and Mandelbrot Set Computation. It is recommended that the global scheduler adopts Guided Self-Scheduling (GSS) for all, and the local scheduler adopts the static scheme for applications with regular workload distribution but any scheme for applications with irregular workload distribution. Experimental results show the best speedups obtained by ELSS for the three benchmark programs are 1.373, 13.34 and 2.4, respectively, compared with that scheduled by LSS.",
author = "Chao-Chin Wu and Lien-Fu Lai and Huang, {Liang Tsung} and Minglung Chen",
year = "2012",
month = "10",
day = "1",
doi = "10.1007/s11227-011-0726-x",
language = "English",
volume = "62",
pages = "399--430",
journal = "Journal of Supercomputing",
issn = "0920-8542",
publisher = "Springer Netherlands",
number = "1",

}

Performance evaluation of enhancement of the layered self-scheduling approach for heterogeneous multicore cluster systems. / Wu, Chao-Chin; Lai, Lien-Fu; Huang, Liang Tsung; Chen, Minglung.

In: Journal of Supercomputing, Vol. 62, No. 1, 01.10.2012, p. 399-430.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Performance evaluation of enhancement of the layered self-scheduling approach for heterogeneous multicore cluster systems

AU - Wu, Chao-Chin

AU - Lai, Lien-Fu

AU - Huang, Liang Tsung

AU - Chen, Minglung

PY - 2012/10/1

Y1 - 2012/10/1

N2 - Previously we have proposed a Layered Self-Scheduling (LSS) approach that is a hybrid MPI and OpenMP based loop self-scheduling approach for dealing with the heterogeneity problem on a cluster system consisting of multi-core compute nodes, where the allocation functions of several well-known schemes have been modified for better performance. Though LSS provides better performance than the conventional self-scheduling schemes, we found the performance can be improved further after our comprehensive experiments and analyses. The newly proposed task scheduling strategy, called Enhanced Layered Self-Scheduling (ELSS), aims at how to utilize the compute powers of multiple processor cores more efficiently in the master compute node and how to schedule tasks to have more stable performance improvements. We have evaluated the new task scheduling strategy by three benchmark applications: Matrix Multiplication, Monte Carlo Integration, and Mandelbrot Set Computation. It is recommended that the global scheduler adopts Guided Self-Scheduling (GSS) for all, and the local scheduler adopts the static scheme for applications with regular workload distribution but any scheme for applications with irregular workload distribution. Experimental results show the best speedups obtained by ELSS for the three benchmark programs are 1.373, 13.34 and 2.4, respectively, compared with that scheduled by LSS.

AB - Previously we have proposed a Layered Self-Scheduling (LSS) approach that is a hybrid MPI and OpenMP based loop self-scheduling approach for dealing with the heterogeneity problem on a cluster system consisting of multi-core compute nodes, where the allocation functions of several well-known schemes have been modified for better performance. Though LSS provides better performance than the conventional self-scheduling schemes, we found the performance can be improved further after our comprehensive experiments and analyses. The newly proposed task scheduling strategy, called Enhanced Layered Self-Scheduling (ELSS), aims at how to utilize the compute powers of multiple processor cores more efficiently in the master compute node and how to schedule tasks to have more stable performance improvements. We have evaluated the new task scheduling strategy by three benchmark applications: Matrix Multiplication, Monte Carlo Integration, and Mandelbrot Set Computation. It is recommended that the global scheduler adopts Guided Self-Scheduling (GSS) for all, and the local scheduler adopts the static scheme for applications with regular workload distribution but any scheme for applications with irregular workload distribution. Experimental results show the best speedups obtained by ELSS for the three benchmark programs are 1.373, 13.34 and 2.4, respectively, compared with that scheduled by LSS.

UR - http://www.scopus.com/inward/record.url?scp=84865614916&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865614916&partnerID=8YFLogxK

U2 - 10.1007/s11227-011-0726-x

DO - 10.1007/s11227-011-0726-x

M3 - Article

AN - SCOPUS:84865614916

VL - 62

SP - 399

EP - 430

JO - Journal of Supercomputing

JF - Journal of Supercomputing

SN - 0920-8542

IS - 1

ER -