A multi-GPU implementation of a D2Q37 lattice Boltzmann code

L. Biferale; F. Mantovani; M. Pivanti; F. Pozzati; M. Sbragaglia; Andrea Scagliarini; S.F. Schifano; F. Toschi; R. Tripiccione

doi:10.1007/978-3-642-31464-3_65

A multi-GPU implementation of a D2Q37 lattice Boltzmann code

L. Biferale, F. Mantovani, M. Pivanti, F. Pozzati, M. Sbragaglia, Andrea Scagliarini, S.F. Schifano, F. Toschi, R. Tripiccione

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

18 Citations (Scopus)

1 Downloads (Pure)

Abstract

We describe a parallel implementation of a compressible Lattice Boltzmann code on a multi-GPU cluster based on Nvidia Fermi processors. We analyze how to optimize the algorithm for GP-GPU architectures, describe the implementation choices that we have adopted and compare our performance results with an implementation optimized for latest generation multi-core CPUs. Our program runs at ˜¿30% of the double-precision peak performance of one GPU and shows almost linear scaling when run on the multi-GPU cluster. Keywords: Computational fluid-dynamics – Lattice Boltzmann methods – GP-GPUs computing

Original language	English
Title of host publication	Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I
Editors	R. Wyrzykowski, J. Dongarra, K. Karczewski, J. Wasniewski
Place of Publication	Berlin
Publisher	Springer
Pages	640-650
ISBN (Print)	978-3-642-31463-6
DOIs	https://doi.org/10.1007/978-3-642-31464-3_65
Publication status	Published - 2012
Event	9th International Conference on Parallel Processing and Applied Mathematics (PPAM 2011) - Torun, Poland Duration: 11 Sep 2011 → 14 Sep 2011 Conference number: 9

Publication series

Name	Lecture Notes in Computer Science
Volume	7203
ISSN (Print)	0302-9743

Conference

Conference	9th International Conference on Parallel Processing and Applied Mathematics (PPAM 2011)
Abbreviated title	PPAM 2011
Country	Poland
City	Torun
Period	11/09/11 → 14/09/11

Access to Document

10.1007/978-3-642-31464-3_65

Fingerprint Dive into the research topics of 'A multi-GPU implementation of a D2Q37 lattice Boltzmann code'. Together they form a unique fingerprint.

Cite this

Biferale, L., Mantovani, F., Pivanti, M., Pozzati, F., Sbragaglia, M., Scagliarini, A., Schifano, S. F., Toschi, F., & Tripiccione, R. (2012). A multi-GPU implementation of a D2Q37 lattice Boltzmann code. In R. Wyrzykowski, J. Dongarra, K. Karczewski, & J. Wasniewski (Eds.), Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I (pp. 640-650). (Lecture Notes in Computer Science; Vol. 7203). Springer. https://doi.org/10.1007/978-3-642-31464-3_65

Biferale, L. ; Mantovani, F. ; Pivanti, M. ; Pozzati, F. ; Sbragaglia, M. ; Scagliarini, Andrea ; Schifano, S.F. ; Toschi, F. ; Tripiccione, R. / A multi-GPU implementation of a D2Q37 lattice Boltzmann code. Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I. editor / R. Wyrzykowski ; J. Dongarra ; K. Karczewski ; J. Wasniewski. Berlin : Springer, 2012. pp. 640-650 (Lecture Notes in Computer Science).

@inproceedings{f815fd2e7520463c9b3833a3cb03f01e,

title = "A multi-GPU implementation of a D2Q37 lattice Boltzmann code",

abstract = "We describe a parallel implementation of a compressible Lattice Boltzmann code on a multi-GPU cluster based on Nvidia Fermi processors. We analyze how to optimize the algorithm for GP-GPU architectures, describe the implementation choices that we have adopted and compare our performance results with an implementation optimized for latest generation multi-core CPUs. Our program runs at ˜¿30% of the double-precision peak performance of one GPU and shows almost linear scaling when run on the multi-GPU cluster. Keywords: Computational fluid-dynamics – Lattice Boltzmann methods – GP-GPUs computing",

author = "L. Biferale and F. Mantovani and M. Pivanti and F. Pozzati and M. Sbragaglia and Andrea Scagliarini and S.F. Schifano and F. Toschi and R. Tripiccione",

year = "2012",

doi = "10.1007/978-3-642-31464-3_65",

language = "English",

isbn = "978-3-642-31463-6",

series = "Lecture Notes in Computer Science",

publisher = "Springer",

pages = "640--650",

editor = "R. Wyrzykowski and J. Dongarra and K. Karczewski and J. Wasniewski",

booktitle = "Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I",

address = "Germany",

note = "9th International Conference on Parallel Processing and Applied Mathematics (PPAM 2011), PPAM 2011 ; Conference date: 11-09-2011 Through 14-09-2011",

}

Biferale, L, Mantovani, F, Pivanti, M, Pozzati, F, Sbragaglia, M, Scagliarini, A, Schifano, SF, Toschi, F & Tripiccione, R 2012, A multi-GPU implementation of a D2Q37 lattice Boltzmann code. in R Wyrzykowski, J Dongarra, K Karczewski & J Wasniewski (eds), Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I. Lecture Notes in Computer Science, vol. 7203, Springer, Berlin, pp. 640-650, 9th International Conference on Parallel Processing and Applied Mathematics (PPAM 2011), Torun, Poland, 11/09/11. https://doi.org/10.1007/978-3-642-31464-3_65

A multi-GPU implementation of a D2Q37 lattice Boltzmann code. / Biferale, L.; Mantovani, F.; Pivanti, M.; Pozzati, F.; Sbragaglia, M.; Scagliarini, Andrea; Schifano, S.F.; Toschi, F.; Tripiccione, R.

Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I. ed. / R. Wyrzykowski; J. Dongarra; K. Karczewski; J. Wasniewski. Berlin : Springer, 2012. p. 640-650 (Lecture Notes in Computer Science; Vol. 7203).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › Academic › peer-review

TY - GEN

T1 - A multi-GPU implementation of a D2Q37 lattice Boltzmann code

AU - Biferale, L.

AU - Mantovani, F.

AU - Pivanti, M.

AU - Pozzati, F.

AU - Sbragaglia, M.

AU - Scagliarini, Andrea

AU - Schifano, S.F.

AU - Toschi, F.

AU - Tripiccione, R.

N1 - Conference code: 9

PY - 2012

Y1 - 2012

N2 - We describe a parallel implementation of a compressible Lattice Boltzmann code on a multi-GPU cluster based on Nvidia Fermi processors. We analyze how to optimize the algorithm for GP-GPU architectures, describe the implementation choices that we have adopted and compare our performance results with an implementation optimized for latest generation multi-core CPUs. Our program runs at ˜¿30% of the double-precision peak performance of one GPU and shows almost linear scaling when run on the multi-GPU cluster. Keywords: Computational fluid-dynamics – Lattice Boltzmann methods – GP-GPUs computing

AB - We describe a parallel implementation of a compressible Lattice Boltzmann code on a multi-GPU cluster based on Nvidia Fermi processors. We analyze how to optimize the algorithm for GP-GPU architectures, describe the implementation choices that we have adopted and compare our performance results with an implementation optimized for latest generation multi-core CPUs. Our program runs at ˜¿30% of the double-precision peak performance of one GPU and shows almost linear scaling when run on the multi-GPU cluster. Keywords: Computational fluid-dynamics – Lattice Boltzmann methods – GP-GPUs computing

U2 - 10.1007/978-3-642-31464-3_65

DO - 10.1007/978-3-642-31464-3_65

M3 - Conference contribution

SN - 978-3-642-31463-6

T3 - Lecture Notes in Computer Science

SP - 640

EP - 650

BT - Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I

A2 - Wyrzykowski, R.

A2 - Dongarra, J.

A2 - Karczewski, K.

A2 - Wasniewski, J.

PB - Springer

CY - Berlin

T2 - 9th International Conference on Parallel Processing and Applied Mathematics (PPAM 2011)

Y2 - 11 September 2011 through 14 September 2011

ER -

Biferale L, Mantovani F, Pivanti M, Pozzati F, Sbragaglia M, Scagliarini A et al. A multi-GPU implementation of a D2Q37 lattice Boltzmann code. In Wyrzykowski R, Dongarra J, Karczewski K, Wasniewski J, editors, Parallel Processing and Applied Mathematics : 9th International Conference, PPAM 2011, Torun, Poland, September 11-14, 2011. Revised Selected Papers, Part I. Berlin: Springer. 2012. p. 640-650. (Lecture Notes in Computer Science). https://doi.org/10.1007/978-3-642-31464-3_65