Adaptive energy minimization of OpenMP parallel applications on many-core systems
Adaptive energy minimization of OpenMP parallel applications on many-core systems
Energy minimization of parallel applications is an emerging challenge for current and future generations of many-core computing systems. In this paper, we propose a novel and scalable energy minimization approach that suitably applies DVFS in the sequential part and jointly considers DVFS and dynamic core allocations in the parallel part. Fundamental to this approach is an iterative learning based control algorithm that adapt the voltage/frequency scaling and core allocations dynamically based on workload predictions and is guided by the CPU performance counters at regular intervals. The adaptation is facilitated through performance annotations in the application codes, defined in a modified OpenMP runtime library. The proposed approach is validated on an Intel Xeon E5-2630 platform with up to 24 CPUs running NAS parallel benchmark applications. We show that our proposed approach can effectively adapt to different architecture and core allocations and minimize energy consumption by up to 17% compared to the existing approaches for a given performance requirement.
many-core, OpenMP, energy minimization
Shafik, Rishad Ahmed
aa0bdafc-b022-4cb2-a8ef-4bf8a03ba524
Das, Anup K.
2a0d6cea-309b-4053-a62e-234807f89306
Yang, Sheng
04b9848f-ddd4-4d8f-93b6-b91a2144d49c
Merrett, Geoff V.
89b3a696-41de-44c3-89aa-b0aa29f54020
Al-Hashimi, Bashir
0b29c671-a6d2-459c-af68-c4614dce3b5d
Shafik, Rishad Ahmed
aa0bdafc-b022-4cb2-a8ef-4bf8a03ba524
Das, Anup K.
2a0d6cea-309b-4053-a62e-234807f89306
Yang, Sheng
04b9848f-ddd4-4d8f-93b6-b91a2144d49c
Merrett, Geoff V.
89b3a696-41de-44c3-89aa-b0aa29f54020
Al-Hashimi, Bashir
0b29c671-a6d2-459c-af68-c4614dce3b5d
Shafik, Rishad Ahmed, Das, Anup K., Yang, Sheng, Merrett, Geoff V. and Al-Hashimi, Bashir
(2015)
Adaptive energy minimization of OpenMP parallel applications on many-core systems.
6th Workshop on Parallel Programming and Run-Time Management Techniques for Many-core Architectures, Amsterdam, Netherlands.
21 Jan 2015.
6 pp
.
Record type:
Conference or Workshop Item
(Paper)
Abstract
Energy minimization of parallel applications is an emerging challenge for current and future generations of many-core computing systems. In this paper, we propose a novel and scalable energy minimization approach that suitably applies DVFS in the sequential part and jointly considers DVFS and dynamic core allocations in the parallel part. Fundamental to this approach is an iterative learning based control algorithm that adapt the voltage/frequency scaling and core allocations dynamically based on workload predictions and is guided by the CPU performance counters at regular intervals. The adaptation is facilitated through performance annotations in the application codes, defined in a modified OpenMP runtime library. The proposed approach is validated on an Intel Xeon E5-2630 platform with up to 24 CPUs running NAS parallel benchmark applications. We show that our proposed approach can effectively adapt to different architecture and core allocations and minimize energy consumption by up to 17% compared to the existing approaches for a given performance requirement.
Text
aspdac2015.pdf
- Accepted Manuscript
More information
e-pub ahead of print date: 2015
Venue - Dates:
6th Workshop on Parallel Programming and Run-Time Management Techniques for Many-core Architectures, Amsterdam, Netherlands, 2015-01-21 - 2015-01-21
Keywords:
many-core, OpenMP, energy minimization
Organisations:
Electronic & Software Systems
Identifiers
Local EPrints ID: 372773
URI: http://eprints.soton.ac.uk/id/eprint/372773
PURE UUID: 22806f7c-7479-4c56-9694-19eace4ee9a7
Catalogue record
Date deposited: 18 Dec 2014 14:47
Last modified: 15 Mar 2024 03:23
Export record
Contributors
Author:
Rishad Ahmed Shafik
Author:
Anup K. Das
Author:
Sheng Yang
Author:
Geoff V. Merrett
Author:
Bashir Al-Hashimi
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics