Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications

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

  • 6 Citations

Abstract

Execution time is no longer the only performance metric for computer systems. In fact, a trend is emerging to trade raw performance for energy savings. Techniques like Dynamic Power Management (DPM, switching to low power state) and Dynamic Voltage and Frequency Scaling (DVFS, throttling processor frequency) help modern systems to reduce their power consumption while adhering to performance requirements. To balance flexibility and design complexity, the concept of Voltage and Frequency Islands (VFIs) was recently introduced for power optimisation. It achieves fine-grained system-level power management, by operating all processors in the same VFI at a common frequency/voltage. This paper presents a novel approach to compute a power management strategy combining DPM and DVFS. In our approach, applications (modelled in full synchronous dataflow, SDF) are mapped on heterogeneous multiprocessor platforms (partitioned in voltage and frequency islands). We compute an energy-optimal schedule, meeting minimal throughput requirements. We demonstrate that the combination of DPM and DVFS provides an energy reduction beyond considering DVFS or DMP separately. Moreover, we show that by clustering processors in VFIs, DPM can be combined with any granularity of DVFS. Our approach uses model checking, by encoding the optimisation problem as a query over priced timed automata. The model-checker UPPAAL Cora extracts a cost minimal trace, representing a power minimal schedule. We illustrate our approach with several case studies on commercially available hardware.
LanguageUndefined
Title of host publication18th Euromicro Conference on Digital Systems Design (DSD 2015)
Place of PublicationUSA
PublisherIEEE Computer Society
Pages271-275
Number of pages5
ISBN (Print)978-1-4673-8035-5
DOIs
StatePublished - 26 Aug 2015
Event18th EUROMICRO Conference on Digital System Design, DSD 2015 - Funchal, Portugal
Duration: 26 Aug 201528 Aug 2015
Conference number: 18

Publication series

Name
PublisherIEEE Computer Society

Conference

Conference18th EUROMICRO Conference on Digital System Design, DSD 2015
Abbreviated titleDSD
CountryPortugal
CityFunchal
Period26/08/1528/08/15

Keywords

  • Priced Timed Automata
  • Power Management
  • Power Optimisation
  • EWI-26398
  • EC Grant Agreement nr.: FP7/2007-2013
  • EC Grant Agreement nr.: FP7/318490
  • METIS-314993
  • Scheduling
  • Processors
  • Voltage and Frequency Scaling
  • IR-98197
  • MPEG-4 Decoder
  • Heterogeneous

Cite this

Ahmad, W., Holzenspies, P. K. F., Stoelinga, M. I. A., & van de Pol, J. C. (2015). Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications. In 18th Euromicro Conference on Digital Systems Design (DSD 2015) (pp. 271-275). USA: IEEE Computer Society. DOI: 10.1109/DSD.2015.59
Ahmad, W. ; Holzenspies, P.K.F. ; Stoelinga, Mariëlle Ida Antoinette ; van de Pol, Jan Cornelis. / Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications. 18th Euromicro Conference on Digital Systems Design (DSD 2015). USA : IEEE Computer Society, 2015. pp. 271-275
@inproceedings{0d747d82dd8f4f9e8d0f7b3ef21ffd2a,
title = "Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications",
abstract = "Execution time is no longer the only performance metric for computer systems. In fact, a trend is emerging to trade raw performance for energy savings. Techniques like Dynamic Power Management (DPM, switching to low power state) and Dynamic Voltage and Frequency Scaling (DVFS, throttling processor frequency) help modern systems to reduce their power consumption while adhering to performance requirements. To balance flexibility and design complexity, the concept of Voltage and Frequency Islands (VFIs) was recently introduced for power optimisation. It achieves fine-grained system-level power management, by operating all processors in the same VFI at a common frequency/voltage. This paper presents a novel approach to compute a power management strategy combining DPM and DVFS. In our approach, applications (modelled in full synchronous dataflow, SDF) are mapped on heterogeneous multiprocessor platforms (partitioned in voltage and frequency islands). We compute an energy-optimal schedule, meeting minimal throughput requirements. We demonstrate that the combination of DPM and DVFS provides an energy reduction beyond considering DVFS or DMP separately. Moreover, we show that by clustering processors in VFIs, DPM can be combined with any granularity of DVFS. Our approach uses model checking, by encoding the optimisation problem as a query over priced timed automata. The model-checker UPPAAL Cora extracts a cost minimal trace, representing a power minimal schedule. We illustrate our approach with several case studies on commercially available hardware.",
keywords = "Priced Timed Automata, Power Management, Power Optimisation, EWI-26398, EC Grant Agreement nr.: FP7/2007-2013, EC Grant Agreement nr.: FP7/318490, METIS-314993, Scheduling, Processors, Voltage and Frequency Scaling, IR-98197, MPEG-4 Decoder, Heterogeneous",
author = "W. Ahmad and P.K.F. Holzenspies and Stoelinga, {Mari{\"e}lle Ida Antoinette} and {van de Pol}, {Jan Cornelis}",
note = "eemcs-eprint-26398",
year = "2015",
month = "8",
day = "26",
doi = "10.1109/DSD.2015.59",
language = "Undefined",
isbn = "978-1-4673-8035-5",
publisher = "IEEE Computer Society",
pages = "271--275",
booktitle = "18th Euromicro Conference on Digital Systems Design (DSD 2015)",
address = "United States",

}

Ahmad, W, Holzenspies, PKF, Stoelinga, MIA & van de Pol, JC 2015, Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications. in 18th Euromicro Conference on Digital Systems Design (DSD 2015). IEEE Computer Society, USA, pp. 271-275, 18th EUROMICRO Conference on Digital System Design, DSD 2015, Funchal, Portugal, 26/08/15. DOI: 10.1109/DSD.2015.59

Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications. / Ahmad, W.; Holzenspies, P.K.F.; Stoelinga, Mariëlle Ida Antoinette; van de Pol, Jan Cornelis.

18th Euromicro Conference on Digital Systems Design (DSD 2015). USA : IEEE Computer Society, 2015. p. 271-275.

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

TY - GEN

T1 - Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications

AU - Ahmad,W.

AU - Holzenspies,P.K.F.

AU - Stoelinga,Mariëlle Ida Antoinette

AU - van de Pol,Jan Cornelis

N1 - eemcs-eprint-26398

PY - 2015/8/26

Y1 - 2015/8/26

N2 - Execution time is no longer the only performance metric for computer systems. In fact, a trend is emerging to trade raw performance for energy savings. Techniques like Dynamic Power Management (DPM, switching to low power state) and Dynamic Voltage and Frequency Scaling (DVFS, throttling processor frequency) help modern systems to reduce their power consumption while adhering to performance requirements. To balance flexibility and design complexity, the concept of Voltage and Frequency Islands (VFIs) was recently introduced for power optimisation. It achieves fine-grained system-level power management, by operating all processors in the same VFI at a common frequency/voltage. This paper presents a novel approach to compute a power management strategy combining DPM and DVFS. In our approach, applications (modelled in full synchronous dataflow, SDF) are mapped on heterogeneous multiprocessor platforms (partitioned in voltage and frequency islands). We compute an energy-optimal schedule, meeting minimal throughput requirements. We demonstrate that the combination of DPM and DVFS provides an energy reduction beyond considering DVFS or DMP separately. Moreover, we show that by clustering processors in VFIs, DPM can be combined with any granularity of DVFS. Our approach uses model checking, by encoding the optimisation problem as a query over priced timed automata. The model-checker UPPAAL Cora extracts a cost minimal trace, representing a power minimal schedule. We illustrate our approach with several case studies on commercially available hardware.

AB - Execution time is no longer the only performance metric for computer systems. In fact, a trend is emerging to trade raw performance for energy savings. Techniques like Dynamic Power Management (DPM, switching to low power state) and Dynamic Voltage and Frequency Scaling (DVFS, throttling processor frequency) help modern systems to reduce their power consumption while adhering to performance requirements. To balance flexibility and design complexity, the concept of Voltage and Frequency Islands (VFIs) was recently introduced for power optimisation. It achieves fine-grained system-level power management, by operating all processors in the same VFI at a common frequency/voltage. This paper presents a novel approach to compute a power management strategy combining DPM and DVFS. In our approach, applications (modelled in full synchronous dataflow, SDF) are mapped on heterogeneous multiprocessor platforms (partitioned in voltage and frequency islands). We compute an energy-optimal schedule, meeting minimal throughput requirements. We demonstrate that the combination of DPM and DVFS provides an energy reduction beyond considering DVFS or DMP separately. Moreover, we show that by clustering processors in VFIs, DPM can be combined with any granularity of DVFS. Our approach uses model checking, by encoding the optimisation problem as a query over priced timed automata. The model-checker UPPAAL Cora extracts a cost minimal trace, representing a power minimal schedule. We illustrate our approach with several case studies on commercially available hardware.

KW - Priced Timed Automata

KW - Power Management

KW - Power Optimisation

KW - EWI-26398

KW - EC Grant Agreement nr.: FP7/2007-2013

KW - EC Grant Agreement nr.: FP7/318490

KW - METIS-314993

KW - Scheduling

KW - Processors

KW - Voltage and Frequency Scaling

KW - IR-98197

KW - MPEG-4 Decoder

KW - Heterogeneous

U2 - 10.1109/DSD.2015.59

DO - 10.1109/DSD.2015.59

M3 - Conference contribution

SN - 978-1-4673-8035-5

SP - 271

EP - 275

BT - 18th Euromicro Conference on Digital Systems Design (DSD 2015)

PB - IEEE Computer Society

CY - USA

ER -

Ahmad W, Holzenspies PKF, Stoelinga MIA, van de Pol JC. Green computing: power optimisation of vfi-based real-time multiprocessor dataflow applications. In 18th Euromicro Conference on Digital Systems Design (DSD 2015). USA: IEEE Computer Society. 2015. p. 271-275. Available from, DOI: 10.1109/DSD.2015.59