A Multi-GPU implementation of a d2q37 lattice boltzmann code

Authors:
Luca Biferale;Filippo Mantovani;Marcello Pivanti;Fabio Pozzati;Mauro Sbragaglia;Andrea Scagliarini;Sebastiano Fabio Schifano;Federico Toschi;Raffaele Tripiccione
Affiliations:
University of Tor Vergata and INFN, Roma, Italy;Deutsches Elektronen Synchrotron (DESY), Zeuthen, Germany;University of Ferrara and INFN, Ferrara, Italy;Fondazione Bruno Kessler Trento, Trento, Italy;University of Tor Vergata and INFN, Roma, Italy;University of Barcelona, Barcelona, Spain;University of Ferrara and INFN, Ferrara, Italy;Eindhoven University of Technology, The Netherlands, CNR-IAC, Rome, Italy;University of Ferrara and INFN, Ferrara, Italy
Venue:
PPAM'11 Proceedings of the 9th international conference on Parallel Processing and Applied Mathematics - Volume Part I
Year:
2011

Citing 4
Cited 0

Performance evaluation of a parallel sparse lattice Boltzmann solver

Journal of Computational Physics
TeraFLOP computing on a desktop PC with GPUs for 3D CFD

International Journal of Computational Fluid Dynamics - Mesoscopic Methods And Their Applications To CFD
Implementation of a Lattice Boltzmann kernel using the Compute Unified Device Architecture developed by nVIDIA

Computing and Visualization in Science
Lattice Boltzmann method simulations on massively parallel multi-core architectures

Proceedings of the 19th High Performance Computing Symposia

Quantified Score

Hi-index	0.00

Visualization

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.