Survey on Efficient Linear Solvers for Porous Media Flow Models on Recent Hardware Architectures. - IFPEN - IFP Energies nouvelles Access content directly
Journal Articles Oil & Gas Science and Technology - Revue d'IFP Energies nouvelles Year : 2014

Survey on Efficient Linear Solvers for Porous Media Flow Models on Recent Hardware Architectures.

Abstract

In the past fewyears,High PerformanceComputing (HPC)technologies led toGeneral Purpose Processing on Graphics Processing Units (GPGPU) andmany-core architectures. These emerging technologies offer massive processing units and are interesting for porous media flow simulators may used forCO2 geological sequestration or EnhancedOil Recovery (EOR) simulation.However the crucial point is "are current algorithms and software able to use these new technologies efficiently?" The resolution of large sparse linear systems, almost ill-conditioned, constitutes the most CPUconsuming part of such simulators. This paper proposes a survey on various solver and preconditioner algorithms, analyzes their efficiency and performance regarding these distinct architectures. Furthermore it proposes a novel approach based on a hybrid programming model for both GPU and many-core clusters. The proposed optimization techniques are validated through a Krylov subspace solver; BiCGStab and some preconditioners like ILU0 on GPU, multi-core and many-core architectures, on various large real study cases in EOR simulation.
Fichier principal
Vignette du fichier
A9RD6C2.pdf (322.73 Ko) Télécharger le fichier
Origin : Publisher files allowed on an open archive
Loading...

Dates and versions

hal-01067895 , version 1 (24-09-2014)

Identifiers

Cite

Ani Anciaux-Sedrakian, Peter Gottschling, Jean-Marc Gratien, Thomas Guignon. Survey on Efficient Linear Solvers for Porous Media Flow Models on Recent Hardware Architectures.. Oil & Gas Science and Technology - Revue d'IFP Energies nouvelles, 2014, 69 (4), pp. 753-766. ⟨10.2516/ogst/2013184⟩. ⟨hal-01067895⟩

Collections

IFP OGST
150 View
490 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More