A Multi-GPU Implementation of a D2Q37 Lattice Boltzmann Code

被引：0

作者：

Biferale, Luca ^{[1
]}

Mantovani, Filippo ^{[2
]}

Pivanti, Marcello ^{[3
,4
]}

Pozzati, Fabio ^{[5
]}

Sbragaglia, Mauro ^{[1
]}

Scagliarini, Andrea ^{[6
]}

Schifano, Sebastiano Fabio ^{[3
,4
]}

Toschi, Federico ^{[7
,8
]}

Tripiccione, Raffaele ^{[3
,4
]}

机构：

[1] Univ Roma Tor Vergata, Rome, Italy

[2] Deutsch Elekt Synchrotron DESY, Zeuthen, Germany

[3] Univ Ferrara, Ferrara, Italy

[4] INFN, Ferrara, Italy

[5] Fondaz Bruno Kessler Trento, Trento, Italy

[6] Univ Barcelona, Barcelona, Spain

[7] Eindhoven Univ Technol, Eindhoven, Netherlands

[8] CNR, IAC, Rome, Italy

来源：

PARALLEL PROCESSING AND APPLIED MATHEMATICS, PT I | 2012年 / 7203卷

关键词：

Computational fluid-dynamics; Lattice Boltzmann methods; GP-GPUs computing; PERFORMANCE;

D O I：

暂无

中图分类号：

TP3 [计算技术、计算机技术];

学科分类号：

0812 ;

摘要：

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 approximate to 30% of the double-precision peak performance of one GPU and shows almost linear scaling when run on the multi-GPU cluster.

引用

页码：640 / 650

页数：11

共 50 条

[1] An optimized D2Q37 Lattice Boltzmann code on GP-GPUs
Biferale, Luca
Mantovani, Filippo
Pivanti, Marcello
Pozzati, Fabio
Sbragaglia, Mauro
Scagliarini, Andrea
Schifano, Sebastiano Fabio
Toschi, Federico
Tripiccione, Raffaele
[J]. COMPUTERS & FLUIDS, 2013, 80 : 55 - 62
[2] An optimized D2Q37 Lattice Boltzmann code on GP-GPUs
[J]. Schifano, S.F. (schifano@fe.infn.it), 2013, Elsevier Ltd (80):
[3] Multi-GPU implementation of the lattice Boltzmann method
Obrecht, Christian
Kuznik, Frederic
Tourancheau, Bernard
Roux, Jean-Jacques
[J]. COMPUTERS & MATHEMATICS WITH APPLICATIONS, 2013, 65 (02) : 252 - 261
[4] Sailfish: A flexible multi-GPU implementation of the lattice Boltzmann method
Januszewski, M.
Kostur, M.
[J]. COMPUTER PHYSICS COMMUNICATIONS, 2014, 185 (09) : 2350 - 2368
[5] The TheLMA project: Multi-GPU implementation of the lattice Boltzmann method
Obrecht, Christian
Kuznik, Frederic
Tourancheau, Bernard
Roux, Jean-Jacques
[J]. INTERNATIONAL JOURNAL OF HIGH PERFORMANCE COMPUTING APPLICATIONS, 2011, 25 (03): : 295 - 303
[6] Optimizing Communications in multi-GPU Lattice Boltzmann Simulations
Calore, Enrico
Marchi, Davide
Schifano, Sebastiano Fabio
Tripiccione, Raffaele
[J]. PROCEEDINGS OF THE 2015 INTERNATIONAL CONFERENCE ON HIGH PERFORMANCE COMPUTING & SIMULATION (HPCS 2015), 2015, : 55 - 62
[7] Implementation of Multi-GPU Based Lattice Boltzmann Method for Flow Through Porous Media
Huang, Changsheng
Shi, Baochang
He, Nanzhong
Chai, Zhenhua
[J]. ADVANCES IN APPLIED MATHEMATICS AND MECHANICS, 2015, 7 (01) : 1 - 12
[8] Multi-GPU implementation of a hybrid thermal lattice Boltzmann solver using the TheLMA framework
Obrecht, Christian
Kuznik, Frederic
Tourancheau, Bernard
Roux, Jean-Jacques
[J]. COMPUTERS & FLUIDS, 2013, 80 : 269 - 275
[9] Adjoint Lattice Boltzmann for topology optimization on multi-GPU architecture
Laniewski-Wollk, L.
Rokicki, J.
[J]. COMPUTERS & MATHEMATICS WITH APPLICATIONS, 2016, 71 (03) : 833 - 848
[10] Performance issues on many-core processors: A D2Q37 Lattice Boltzmann scheme as a test-case
Mantovani, F.
Pivanti, M.
Schifano, S. F.
Tripiccione, R.
[J]. COMPUTERS & FLUIDS, 2013, 88 : 743 - 752

← 1 2 3 4 5 →