An ILP formulation for task scheduling on heterogeneous chip multiprocessors

Suleyman Tosun, Nazanin Mansouri, Mahmut Kandemir, Ozcan Ozturk

Research output: Chapter in Book/Report/Conference proceedingConference contribution

7 Citations (Scopus)

Abstract

One of the main difficuties to map an embedded application onto a multiprocessor architecture is that there are multiple ways of this mapping due to several constraints. In this paper, we present an Integer Linear Programming based framework that maps a given application (represented as a task graph) onto a Heterogeneous Chip Multiprocessor architecture. Our framework can be used with several objective functions such as energy, performance, and fallibility (opposite of reliability). We use Dynamic Voltage Scaling (DVS) for reducing energy consumption while we employ task duplication to minimize fallibility. Our experimental results show that over 50% improvements on energy consumption are possible by using DVS, and the fully task duplicated schedules can be achieved under tight performance and energy bounds.

Original languageEnglish (US)
Title of host publicationComputer and Information Sciences - ISCIS 2006
Subtitle of host publication21th International Symposium, Proceedings
PublisherSpringer Verlag
Pages267-276
Number of pages10
ISBN (Print)3540472428, 9783540472421
StatePublished - Jan 1 2006
EventISCIS 2006: 21th International Symposium on Computer and Information Sciences - Istanbul, Turkey
Duration: Nov 1 2006Nov 3 2006

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume4263 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

OtherISCIS 2006: 21th International Symposium on Computer and Information Sciences
CountryTurkey
CityIstanbul
Period11/1/0611/3/06

Fingerprint

Inductive logic programming (ILP)
Chip multiprocessors
Task Scheduling
Dynamic Voltage Scaling
Energy utilization
Scheduling
Energy Consumption
Formulation
Linear programming
Task Graph
Integer Linear Programming
Duplication
Multiprocessor
Energy
Schedule
Objective function
Minimise
Experimental Results
Voltage scaling
Framework

All Science Journal Classification (ASJC) codes

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Tosun, S., Mansouri, N., Kandemir, M., & Ozturk, O. (2006). An ILP formulation for task scheduling on heterogeneous chip multiprocessors. In Computer and Information Sciences - ISCIS 2006: 21th International Symposium, Proceedings (pp. 267-276). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4263 LNCS). Springer Verlag.
Tosun, Suleyman ; Mansouri, Nazanin ; Kandemir, Mahmut ; Ozturk, Ozcan. / An ILP formulation for task scheduling on heterogeneous chip multiprocessors. Computer and Information Sciences - ISCIS 2006: 21th International Symposium, Proceedings. Springer Verlag, 2006. pp. 267-276 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{309866c4fe554a0185033e40b7e92d6e,
title = "An ILP formulation for task scheduling on heterogeneous chip multiprocessors",
abstract = "One of the main difficuties to map an embedded application onto a multiprocessor architecture is that there are multiple ways of this mapping due to several constraints. In this paper, we present an Integer Linear Programming based framework that maps a given application (represented as a task graph) onto a Heterogeneous Chip Multiprocessor architecture. Our framework can be used with several objective functions such as energy, performance, and fallibility (opposite of reliability). We use Dynamic Voltage Scaling (DVS) for reducing energy consumption while we employ task duplication to minimize fallibility. Our experimental results show that over 50{\%} improvements on energy consumption are possible by using DVS, and the fully task duplicated schedules can be achieved under tight performance and energy bounds.",
author = "Suleyman Tosun and Nazanin Mansouri and Mahmut Kandemir and Ozcan Ozturk",
year = "2006",
month = "1",
day = "1",
language = "English (US)",
isbn = "3540472428",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "267--276",
booktitle = "Computer and Information Sciences - ISCIS 2006",
address = "Germany",

}

Tosun, S, Mansouri, N, Kandemir, M & Ozturk, O 2006, An ILP formulation for task scheduling on heterogeneous chip multiprocessors. in Computer and Information Sciences - ISCIS 2006: 21th International Symposium, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 4263 LNCS, Springer Verlag, pp. 267-276, ISCIS 2006: 21th International Symposium on Computer and Information Sciences, Istanbul, Turkey, 11/1/06.

An ILP formulation for task scheduling on heterogeneous chip multiprocessors. / Tosun, Suleyman; Mansouri, Nazanin; Kandemir, Mahmut; Ozturk, Ozcan.

Computer and Information Sciences - ISCIS 2006: 21th International Symposium, Proceedings. Springer Verlag, 2006. p. 267-276 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 4263 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - An ILP formulation for task scheduling on heterogeneous chip multiprocessors

AU - Tosun, Suleyman

AU - Mansouri, Nazanin

AU - Kandemir, Mahmut

AU - Ozturk, Ozcan

PY - 2006/1/1

Y1 - 2006/1/1

N2 - One of the main difficuties to map an embedded application onto a multiprocessor architecture is that there are multiple ways of this mapping due to several constraints. In this paper, we present an Integer Linear Programming based framework that maps a given application (represented as a task graph) onto a Heterogeneous Chip Multiprocessor architecture. Our framework can be used with several objective functions such as energy, performance, and fallibility (opposite of reliability). We use Dynamic Voltage Scaling (DVS) for reducing energy consumption while we employ task duplication to minimize fallibility. Our experimental results show that over 50% improvements on energy consumption are possible by using DVS, and the fully task duplicated schedules can be achieved under tight performance and energy bounds.

AB - One of the main difficuties to map an embedded application onto a multiprocessor architecture is that there are multiple ways of this mapping due to several constraints. In this paper, we present an Integer Linear Programming based framework that maps a given application (represented as a task graph) onto a Heterogeneous Chip Multiprocessor architecture. Our framework can be used with several objective functions such as energy, performance, and fallibility (opposite of reliability). We use Dynamic Voltage Scaling (DVS) for reducing energy consumption while we employ task duplication to minimize fallibility. Our experimental results show that over 50% improvements on energy consumption are possible by using DVS, and the fully task duplicated schedules can be achieved under tight performance and energy bounds.

UR - http://www.scopus.com/inward/record.url?scp=33845241669&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33845241669&partnerID=8YFLogxK

M3 - Conference contribution

SN - 3540472428

SN - 9783540472421

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 267

EP - 276

BT - Computer and Information Sciences - ISCIS 2006

PB - Springer Verlag

ER -

Tosun S, Mansouri N, Kandemir M, Ozturk O. An ILP formulation for task scheduling on heterogeneous chip multiprocessors. In Computer and Information Sciences - ISCIS 2006: 21th International Symposium, Proceedings. Springer Verlag. 2006. p. 267-276. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).