Reliable mapping and partitioning of performance-constrained OpenCL Applications on CPU-GPU MPSoCs
Reliable mapping and partitioning of performance-constrained OpenCL Applications on CPU-GPU MPSoCs
Heterogeneous Multi-Processor Systems-on-Chips (MPSoCs) containing CPU and GPU cores are typically required to execute applications concurrently. Existing approaches exploit applications executing in CPU and GPU cores at the same time taking into account performance and energy consumption for mapping and partitioning. This paper presents a proposal for mapping and partitioning of applications in CPU-GPU MPSoCs taking into account the temperature behavior of the system. We evaluate the temperature profiling to partition the applications between CPU and GPU. The profiling is done by measuring the temperature of the CPU and GPU cores while executing different applications at different
partitions. Results shown up to 13% savings of average temperature of the chip while maintaining performance requirements. A lower thermal behavior represents a better long-term reliability (lifetime) of the SoC.
Heterogeneous MPSoC, OpenCL applications, Mapping, Partitioning, Performance, Energy consumption, Temperature-aware
Weber Wachter, Eduardo
bdacc537-b1ac-4241-a6fc-b67f1e6a6ce8
Merrett, Geoff V.
89b3a696-41de-44c3-89aa-b0aa29f54020
Singh, Amit
bb67d43e-34d9-4b58-9295-8b5458270408
Al-Hashimi, Bashir
0b29c671-a6d2-459c-af68-c4614dce3b5d
15 October 2017
Weber Wachter, Eduardo
bdacc537-b1ac-4241-a6fc-b67f1e6a6ce8
Merrett, Geoff V.
89b3a696-41de-44c3-89aa-b0aa29f54020
Singh, Amit
bb67d43e-34d9-4b58-9295-8b5458270408
Al-Hashimi, Bashir
0b29c671-a6d2-459c-af68-c4614dce3b5d
Weber Wachter, Eduardo, Merrett, Geoff V., Singh, Amit and Al-Hashimi, Bashir
(2017)
Reliable mapping and partitioning of performance-constrained OpenCL Applications on CPU-GPU MPSoCs.
15th IEEE/ACM Symposium on Embedded Systems for Real-Time Multimedia, , Seoul, Korea, Republic of.
15 - 20 Oct 2017.
(doi:10.1145/3139315.3157088).
Record type:
Conference or Workshop Item
(Paper)
Abstract
Heterogeneous Multi-Processor Systems-on-Chips (MPSoCs) containing CPU and GPU cores are typically required to execute applications concurrently. Existing approaches exploit applications executing in CPU and GPU cores at the same time taking into account performance and energy consumption for mapping and partitioning. This paper presents a proposal for mapping and partitioning of applications in CPU-GPU MPSoCs taking into account the temperature behavior of the system. We evaluate the temperature profiling to partition the applications between CPU and GPU. The profiling is done by measuring the temperature of the CPU and GPU cores while executing different applications at different
partitions. Results shown up to 13% savings of average temperature of the chip while maintaining performance requirements. A lower thermal behavior represents a better long-term reliability (lifetime) of the SoC.
Text
estimedia_2017
- Accepted Manuscript
More information
Accepted/In Press date: 8 September 2017
e-pub ahead of print date: 15 October 2017
Published date: 15 October 2017
Venue - Dates:
15th IEEE/ACM Symposium on Embedded Systems for Real-Time Multimedia, , Seoul, Korea, Republic of, 2017-10-15 - 2017-10-20
Keywords:
Heterogeneous MPSoC, OpenCL applications, Mapping, Partitioning, Performance, Energy consumption, Temperature-aware
Identifiers
Local EPrints ID: 414757
URI: http://eprints.soton.ac.uk/id/eprint/414757
PURE UUID: cd951d30-9fcf-49e2-acac-dde0ebfd1d61
Catalogue record
Date deposited: 10 Oct 2017 16:31
Last modified: 16 Mar 2024 03:46
Export record
Altmetrics
Contributors
Author:
Eduardo Weber Wachter
Author:
Geoff V. Merrett
Author:
Amit Singh
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