Cost effective implementation of flux limiter functions using partial reconfiguration

Mohamad Sofian Abu Talip, Takayuki Akamine, Yasunori Osana, Naoyuki Fujita, Hideharu Amano

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

1 Citation (Scopus)

Abstract

Computational Fluid Dynamics (CFD) is used as a common design tool in aerospace industry. UPACS, a package for CFD is convenient for users, since a customized simulator can be built just by selecting required functions. The problem is its computation speed which is hard to be enhanced by using clusters due to its complex memory access patterns. As an economical solution, accelerators using FPGAs are hopeful candidates. However, the total scale of UPACS is too large to be implemented on small numbers of FPGAs. For cost efficient implementation, partial reconfiguration which can dynamically reconfigure only required functions is proposed in this paper. Here, MUSCL algorithm used frequently in UPACS is selected as a target. Partial reconfiguration is applied to the flux limiter functions (FLF) in MUSCL. Four FLFs are implemented for Turbulence MUSCL (TMUSCL) and eight FLFs are for Convection MUSCL (CMUSCL). All FLFs are developed independently and separated from the top MUSCL module. At start-up, only required FLFs are selected and deployed to the system without interfering the other modules. This implementation has successfully reduced the resource utilization by 44% to 63%. Total power consumption also reduced by 33%. Configuration speed is improved by 34-times faster as compared to fully reconfiguration method. All implemented functions achieved at least 17 times speed-up compared with the software implementation.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages215-226
Number of pages12
Volume7199 LNCS
DOIs
Publication statusPublished - 2012
Event8th International Symposium on Applied Reconfigurable Computing, ARC 2012 - Hong Kong, China
Duration: 2012 Mar 192012 Mar 23

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7199 LNCS
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other8th International Symposium on Applied Reconfigurable Computing, ARC 2012
CountryChina
CityHong Kong
Period12/3/1912/3/23

Fingerprint

Limiter
Limiters
Reconfiguration
Fluxes
Partial
Costs
Computational Fluid Dynamics
Field Programmable Gate Array
Field programmable gate arrays (FPGA)
Computational fluid dynamics
Module
Aerospace industry
Start-up
Accelerator
Efficient Implementation
Power Consumption
Particle accelerators
Convection
Turbulence
Simulator

Keywords

  • CFD
  • FPGA
  • Partial Reconfiguration
  • Reconfigurable Hardware
  • Scientific Computations

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Abu Talip, M. S., Akamine, T., Osana, Y., Fujita, N., & Amano, H. (2012). Cost effective implementation of flux limiter functions using partial reconfiguration. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7199 LNCS, pp. 215-226). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7199 LNCS). https://doi.org/10.1007/978-3-642-28365-9_18

Cost effective implementation of flux limiter functions using partial reconfiguration. / Abu Talip, Mohamad Sofian; Akamine, Takayuki; Osana, Yasunori; Fujita, Naoyuki; Amano, Hideharu.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7199 LNCS 2012. p. 215-226 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7199 LNCS).

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

Abu Talip, MS, Akamine, T, Osana, Y, Fujita, N & Amano, H 2012, Cost effective implementation of flux limiter functions using partial reconfiguration. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7199 LNCS, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7199 LNCS, pp. 215-226, 8th International Symposium on Applied Reconfigurable Computing, ARC 2012, Hong Kong, China, 12/3/19. https://doi.org/10.1007/978-3-642-28365-9_18
Abu Talip MS, Akamine T, Osana Y, Fujita N, Amano H. Cost effective implementation of flux limiter functions using partial reconfiguration. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7199 LNCS. 2012. p. 215-226. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-28365-9_18
Abu Talip, Mohamad Sofian ; Akamine, Takayuki ; Osana, Yasunori ; Fujita, Naoyuki ; Amano, Hideharu. / Cost effective implementation of flux limiter functions using partial reconfiguration. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7199 LNCS 2012. pp. 215-226 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{e3a2a9bd1a4b414ba091e5398e0cd52a,
title = "Cost effective implementation of flux limiter functions using partial reconfiguration",
abstract = "Computational Fluid Dynamics (CFD) is used as a common design tool in aerospace industry. UPACS, a package for CFD is convenient for users, since a customized simulator can be built just by selecting required functions. The problem is its computation speed which is hard to be enhanced by using clusters due to its complex memory access patterns. As an economical solution, accelerators using FPGAs are hopeful candidates. However, the total scale of UPACS is too large to be implemented on small numbers of FPGAs. For cost efficient implementation, partial reconfiguration which can dynamically reconfigure only required functions is proposed in this paper. Here, MUSCL algorithm used frequently in UPACS is selected as a target. Partial reconfiguration is applied to the flux limiter functions (FLF) in MUSCL. Four FLFs are implemented for Turbulence MUSCL (TMUSCL) and eight FLFs are for Convection MUSCL (CMUSCL). All FLFs are developed independently and separated from the top MUSCL module. At start-up, only required FLFs are selected and deployed to the system without interfering the other modules. This implementation has successfully reduced the resource utilization by 44{\%} to 63{\%}. Total power consumption also reduced by 33{\%}. Configuration speed is improved by 34-times faster as compared to fully reconfiguration method. All implemented functions achieved at least 17 times speed-up compared with the software implementation.",
keywords = "CFD, FPGA, Partial Reconfiguration, Reconfigurable Hardware, Scientific Computations",
author = "{Abu Talip}, {Mohamad Sofian} and Takayuki Akamine and Yasunori Osana and Naoyuki Fujita and Hideharu Amano",
year = "2012",
doi = "10.1007/978-3-642-28365-9_18",
language = "English",
isbn = "9783642283642",
volume = "7199 LNCS",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "215--226",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Cost effective implementation of flux limiter functions using partial reconfiguration

AU - Abu Talip, Mohamad Sofian

AU - Akamine, Takayuki

AU - Osana, Yasunori

AU - Fujita, Naoyuki

AU - Amano, Hideharu

PY - 2012

Y1 - 2012

N2 - Computational Fluid Dynamics (CFD) is used as a common design tool in aerospace industry. UPACS, a package for CFD is convenient for users, since a customized simulator can be built just by selecting required functions. The problem is its computation speed which is hard to be enhanced by using clusters due to its complex memory access patterns. As an economical solution, accelerators using FPGAs are hopeful candidates. However, the total scale of UPACS is too large to be implemented on small numbers of FPGAs. For cost efficient implementation, partial reconfiguration which can dynamically reconfigure only required functions is proposed in this paper. Here, MUSCL algorithm used frequently in UPACS is selected as a target. Partial reconfiguration is applied to the flux limiter functions (FLF) in MUSCL. Four FLFs are implemented for Turbulence MUSCL (TMUSCL) and eight FLFs are for Convection MUSCL (CMUSCL). All FLFs are developed independently and separated from the top MUSCL module. At start-up, only required FLFs are selected and deployed to the system without interfering the other modules. This implementation has successfully reduced the resource utilization by 44% to 63%. Total power consumption also reduced by 33%. Configuration speed is improved by 34-times faster as compared to fully reconfiguration method. All implemented functions achieved at least 17 times speed-up compared with the software implementation.

AB - Computational Fluid Dynamics (CFD) is used as a common design tool in aerospace industry. UPACS, a package for CFD is convenient for users, since a customized simulator can be built just by selecting required functions. The problem is its computation speed which is hard to be enhanced by using clusters due to its complex memory access patterns. As an economical solution, accelerators using FPGAs are hopeful candidates. However, the total scale of UPACS is too large to be implemented on small numbers of FPGAs. For cost efficient implementation, partial reconfiguration which can dynamically reconfigure only required functions is proposed in this paper. Here, MUSCL algorithm used frequently in UPACS is selected as a target. Partial reconfiguration is applied to the flux limiter functions (FLF) in MUSCL. Four FLFs are implemented for Turbulence MUSCL (TMUSCL) and eight FLFs are for Convection MUSCL (CMUSCL). All FLFs are developed independently and separated from the top MUSCL module. At start-up, only required FLFs are selected and deployed to the system without interfering the other modules. This implementation has successfully reduced the resource utilization by 44% to 63%. Total power consumption also reduced by 33%. Configuration speed is improved by 34-times faster as compared to fully reconfiguration method. All implemented functions achieved at least 17 times speed-up compared with the software implementation.

KW - CFD

KW - FPGA

KW - Partial Reconfiguration

KW - Reconfigurable Hardware

KW - Scientific Computations

UR - http://www.scopus.com/inward/record.url?scp=84859463466&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84859463466&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-28365-9_18

DO - 10.1007/978-3-642-28365-9_18

M3 - Conference contribution

SN - 9783642283642

VL - 7199 LNCS

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

SP - 215

EP - 226

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

ER -