GPU acceleration of the caffa3d.MB model

Authors:
Pablo Igounet;Pablo Alfaro;Gabriel Usera;Pablo Ezzatti
Affiliations:
Instituto de Computación, Universidad de la República, Montevideo, Uruguay;Instituto de Computación, Universidad de la República, Montevideo, Uruguay;Instituto de Mecánica de los Fluidos e Ingeniería Ambiental, Universidad de la República, Montevideo, Uruguay;Instituto de Computación, Universidad de la República, Montevideo, Uruguay
Venue:
ICCSA'12 Proceedings of the 12th international conference on Computational Science and Its Applications - Volume Part IV
Year:
2012

Citing 4
Cited 0

Accurate three-dimensional lid-driven cavity flow

Journal of Computational Physics
Exploiting the capabilities of modern GPUs for dense matrix computations

Concurrency and Computation: Practice & Experience
Programming Massively Parallel Processors: A Hands-on Approach

Programming Massively Parallel Processors: A Hands-on Approach
Using hybrid CPU-GPU platforms to accelerate the computation of the matrix sign function

Euro-Par'09 Proceedings of the 2009 international conference on Parallel processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

This work presents a study of porting Strongly Implicit Procedure (SIP) solver to GPU in order to improve its computational efficiency. The SIP heptadiagonal linear system solver was evaluated to be the most time consuming stage in finite volume flow solver caffa3d.MB . The experimental evaluation of the proposed implementation of the solver demonstrates that a significant runtime reduction can be attained (acceleration values up to 10×) when compared with a CPU version, and this improvement significantly reduces the total runtime of the model. This results evidence a promising prospect for a full GPU-based implementation of finite volume flow solvers like caffa3d.MB .