Publication result detail

A GPU solver for symmetric positive-definite matrices vs. traditional codes

BOHÁČEK, J.; KARIMI-SIBAKI, E.; KHARICHA, A.; LUDWIG, A.; WU, M.; HOLZMANN, T.;

Original Title

A GPU solver for symmetric positive-definite matrices vs. traditional codes

English Title

A GPU solver for symmetric positive-definite matrices vs. traditional codes

Type

WoS Article

Original Abstract

In Heat Transfer and Fluid Flow Laboratory in Brno, the inverse heat conduction problem (IHCP) has been extensively used to reconstruct thermal boundary conditions at hot surfaces of solid materials cooled by spraying nozzles. More than three decades of experience and cooperation with industries has proven our experimental/numerical technique to be reliable and very accurate. However, a typical calculation requires relatively long calculation time. The transient heat diffusion in a multi-material sample is the most computationally costly ingredient of the algorithm. In the present paper, the potential for speeding up our calculations is manifested by firstly benchmarking it against traditional CFD codes such as OpenFOAM (FDIC) and ANSYS Fluent (AMG). Secondly, we also unveil a unique comparison between the performance of three inhouse GPU codes each written by a different PhD student/postdoc. Chronologically listed, one student pushed his luck with a fully explicit scheme, while the other two, including us, bet on implicit methods namely the line-by-line method in OpenCL and the conjugate gradient method with the deflated truncated Neumann series preconditioner in CUDA C. (C) 2019 The Authors. Published by Elsevier Ltd.

English abstract

In Heat Transfer and Fluid Flow Laboratory in Brno, the inverse heat conduction problem (IHCP) has been extensively used to reconstruct thermal boundary conditions at hot surfaces of solid materials cooled by spraying nozzles. More than three decades of experience and cooperation with industries has proven our experimental/numerical technique to be reliable and very accurate. However, a typical calculation requires relatively long calculation time. The transient heat diffusion in a multi-material sample is the most computationally costly ingredient of the algorithm. In the present paper, the potential for speeding up our calculations is manifested by firstly benchmarking it against traditional CFD codes such as OpenFOAM (FDIC) and ANSYS Fluent (AMG). Secondly, we also unveil a unique comparison between the performance of three inhouse GPU codes each written by a different PhD student/postdoc. Chronologically listed, one student pushed his luck with a fully explicit scheme, while the other two, including us, bet on implicit methods namely the line-by-line method in OpenCL and the conjugate gradient method with the deflated truncated Neumann series preconditioner in CUDA C. (C) 2019 The Authors. Published by Elsevier Ltd.

Keywords

Linear solver; Inverse task; Heat transfer; GPU; CUDA; OpenFOAM

Key words in English

Linear solver; Inverse task; Heat transfer; GPU; CUDA; OpenFOAM

Authors

BOHÁČEK, J.; KARIMI-SIBAKI, E.; KHARICHA, A.; LUDWIG, A.; WU, M.; HOLZMANN, T.;

RIV year

2020

Released

07.03.2019

Publisher

PERGAMON-ELSEVIER SCIENCE LTD

Location

OXFORD

ISBN

0898-1221

Periodical

COMPUTERS & MATHEMATICS WITH APPLICATIONS

Volume

78

Number

9

State

United Kingdom of Great Britain and Northern Ireland

Pages from

2933

Pages to

2943

Pages count

11

URL

https://www.sciencedirect.com/science/article/pii/S0898122119301105

BibTex

@article{BUT163681,
  author="BOHÁČEK, J. and KARIMI-SIBAKI, E. and KHARICHA, A. and LUDWIG, A. and WU, M. and HOLZMANN, T.",
  title="A GPU solver for symmetric positive-definite matrices vs. traditional codes",
  journal="COMPUTERS & MATHEMATICS WITH APPLICATIONS",
  year="2019",
  volume="78",
  number="9",
  pages="2933--2943",
  doi="10.1016/j.camwa.2019.02.034",
  issn="0898-1221",
  url="https://www.sciencedirect.com/science/article/pii/S0898122119301105"
}