Resource recycling: Putting idle resources to work on a composable accelerator

Yongjun Park, Hyunchul Park, Scott Mahlke, Sukjin Kim

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

2 Citations (Scopus)

Abstract

Mobile computing platforms in the form of smart phones, netbooks, and personal digital assistants have become an integral part of our everyday lives. Moving ahead to the future, mobile multimedia support will become a key differentiating factor for customers. Features such as high-definition audio and video, video conferencing, 3D graphics, and image projection will lead to the adoption of one phone over another. However, in contrast to wireless signal processing which is dominated by vectorizable computation, mobile multimedia applications often contain complex control flow and variable computational requirements. Moreover, data access is more complex where media applications typically operate on multi-dimensional vectors of data rather than single-dimensional vectors with simple strides. To handle these complexities, composable accelerators such as the Polymorphic Pipeline Array, or PPA, present an appealing hardware platform by adding a degree of hardware configurability over existing accelerators. Hardware resources can be both statically as well as dynamically partitioned among executing tasks to maximize execution efficiency. However, an effective compilation framework is essential to partition and assign resources to make intelligent use of the available hardware. In this paper, a compilation framework is introduced that maximizes application throughput with hybrid resource partitioning of a PPA system. Static partitioning handles part of the resource assignment, but this is followed up by dynamic partitioning to identify idle resources and put them to use - resource recycling. Experimental results show that real-time media applications can take advantage of the static and dynamic configurability of the PPA for increased throughput.

Original languageEnglish
Title of host publicationEmbedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10
Pages21-30
Number of pages10
DOIs
StatePublished - 2010 Dec 1
Event6th Embedded Systems Week 2010, ESWEEK 2010 - 2010 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES'10 - Scottsdale, AZ, United States
Duration: 2010 Oct 242010 Oct 29

Other

Other6th Embedded Systems Week 2010, ESWEEK 2010 - 2010 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES'10
CountryUnited States
CityScottsdale, AZ
Period10/10/2410/10/29

Fingerprint

Particle accelerators
Recycling
Hardware
Throughput
Video conferencing
Mobile computing
Personal digital assistants
Flow control
Signal processing
Pipelines

Keywords

  • Coarse-grained reconfigurable architecture
  • Composable accelerator
  • Dynamic partitioning
  • Modulo Scheduling
  • Workload balancing

Cite this

Park, Y., Park, H., Mahlke, S., & Kim, S. (2010). Resource recycling: Putting idle resources to work on a composable accelerator. In Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10 (pp. 21-30) https://doi.org/10.1145/1878921.1878925
Park, Yongjun ; Park, Hyunchul ; Mahlke, Scott ; Kim, Sukjin. / Resource recycling : Putting idle resources to work on a composable accelerator. Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10. 2010. pp. 21-30
@inproceedings{0e9e8ba8586044aeaaadd8df440a6683,
title = "Resource recycling: Putting idle resources to work on a composable accelerator",
abstract = "Mobile computing platforms in the form of smart phones, netbooks, and personal digital assistants have become an integral part of our everyday lives. Moving ahead to the future, mobile multimedia support will become a key differentiating factor for customers. Features such as high-definition audio and video, video conferencing, 3D graphics, and image projection will lead to the adoption of one phone over another. However, in contrast to wireless signal processing which is dominated by vectorizable computation, mobile multimedia applications often contain complex control flow and variable computational requirements. Moreover, data access is more complex where media applications typically operate on multi-dimensional vectors of data rather than single-dimensional vectors with simple strides. To handle these complexities, composable accelerators such as the Polymorphic Pipeline Array, or PPA, present an appealing hardware platform by adding a degree of hardware configurability over existing accelerators. Hardware resources can be both statically as well as dynamically partitioned among executing tasks to maximize execution efficiency. However, an effective compilation framework is essential to partition and assign resources to make intelligent use of the available hardware. In this paper, a compilation framework is introduced that maximizes application throughput with hybrid resource partitioning of a PPA system. Static partitioning handles part of the resource assignment, but this is followed up by dynamic partitioning to identify idle resources and put them to use - resource recycling. Experimental results show that real-time media applications can take advantage of the static and dynamic configurability of the PPA for increased throughput.",
keywords = "Coarse-grained reconfigurable architecture, Composable accelerator, Dynamic partitioning, Modulo Scheduling, Workload balancing",
author = "Yongjun Park and Hyunchul Park and Scott Mahlke and Sukjin Kim",
year = "2010",
month = "12",
day = "1",
doi = "10.1145/1878921.1878925",
language = "English",
isbn = "9781605589039",
pages = "21--30",
booktitle = "Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10",

}

Park, Y, Park, H, Mahlke, S & Kim, S 2010, Resource recycling: Putting idle resources to work on a composable accelerator. in Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10. pp. 21-30, 6th Embedded Systems Week 2010, ESWEEK 2010 - 2010 International Conference on Compilers, Architecture, and Synthesis for Embedded Systems, CASES'10, Scottsdale, AZ, United States, 10/10/24. https://doi.org/10.1145/1878921.1878925

Resource recycling : Putting idle resources to work on a composable accelerator. / Park, Yongjun; Park, Hyunchul; Mahlke, Scott; Kim, Sukjin.

Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10. 2010. p. 21-30.

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

TY - GEN

T1 - Resource recycling

T2 - Putting idle resources to work on a composable accelerator

AU - Park, Yongjun

AU - Park, Hyunchul

AU - Mahlke, Scott

AU - Kim, Sukjin

PY - 2010/12/1

Y1 - 2010/12/1

N2 - Mobile computing platforms in the form of smart phones, netbooks, and personal digital assistants have become an integral part of our everyday lives. Moving ahead to the future, mobile multimedia support will become a key differentiating factor for customers. Features such as high-definition audio and video, video conferencing, 3D graphics, and image projection will lead to the adoption of one phone over another. However, in contrast to wireless signal processing which is dominated by vectorizable computation, mobile multimedia applications often contain complex control flow and variable computational requirements. Moreover, data access is more complex where media applications typically operate on multi-dimensional vectors of data rather than single-dimensional vectors with simple strides. To handle these complexities, composable accelerators such as the Polymorphic Pipeline Array, or PPA, present an appealing hardware platform by adding a degree of hardware configurability over existing accelerators. Hardware resources can be both statically as well as dynamically partitioned among executing tasks to maximize execution efficiency. However, an effective compilation framework is essential to partition and assign resources to make intelligent use of the available hardware. In this paper, a compilation framework is introduced that maximizes application throughput with hybrid resource partitioning of a PPA system. Static partitioning handles part of the resource assignment, but this is followed up by dynamic partitioning to identify idle resources and put them to use - resource recycling. Experimental results show that real-time media applications can take advantage of the static and dynamic configurability of the PPA for increased throughput.

AB - Mobile computing platforms in the form of smart phones, netbooks, and personal digital assistants have become an integral part of our everyday lives. Moving ahead to the future, mobile multimedia support will become a key differentiating factor for customers. Features such as high-definition audio and video, video conferencing, 3D graphics, and image projection will lead to the adoption of one phone over another. However, in contrast to wireless signal processing which is dominated by vectorizable computation, mobile multimedia applications often contain complex control flow and variable computational requirements. Moreover, data access is more complex where media applications typically operate on multi-dimensional vectors of data rather than single-dimensional vectors with simple strides. To handle these complexities, composable accelerators such as the Polymorphic Pipeline Array, or PPA, present an appealing hardware platform by adding a degree of hardware configurability over existing accelerators. Hardware resources can be both statically as well as dynamically partitioned among executing tasks to maximize execution efficiency. However, an effective compilation framework is essential to partition and assign resources to make intelligent use of the available hardware. In this paper, a compilation framework is introduced that maximizes application throughput with hybrid resource partitioning of a PPA system. Static partitioning handles part of the resource assignment, but this is followed up by dynamic partitioning to identify idle resources and put them to use - resource recycling. Experimental results show that real-time media applications can take advantage of the static and dynamic configurability of the PPA for increased throughput.

KW - Coarse-grained reconfigurable architecture

KW - Composable accelerator

KW - Dynamic partitioning

KW - Modulo Scheduling

KW - Workload balancing

UR - http://www.scopus.com/inward/record.url?scp=78650654287&partnerID=8YFLogxK

U2 - 10.1145/1878921.1878925

DO - 10.1145/1878921.1878925

M3 - Conference contribution

AN - SCOPUS:78650654287

SN - 9781605589039

SP - 21

EP - 30

BT - Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10

ER -

Park Y, Park H, Mahlke S, Kim S. Resource recycling: Putting idle resources to work on a composable accelerator. In Embedded Systems Week 2010 - Proceedings of the 2010 International Conference on Compilers, Architecture and Synthesis for Embedded Systems, CASES'10. 2010. p. 21-30 https://doi.org/10.1145/1878921.1878925