Large calculation of the flow over a hypersonic vehicle using a GPU

Elsen, Erich; LeGresley, Patrick; Darve, Eric

doi:10.1016/j.jcp.2008.08.023

Cited by 211 publications

(105 citation statements)

References 18 publications

Supporting

Mentioning

101

Contrasting

Unclassified

Order By: Relevance

“…Then, the ensuing exact solutions are discretized to form a system of algebraic equations that can be solved using an iterative method [15,17], which includes a procedure to ensure the coupling between velocity and pressure fields. At the end, any of the discretized governing equations can be written in the following general linear form at any cell P for the property (u for instance) I ¦ S a a nb nb nb I I P P (7) in which contributions to the central and neighbour cells are carried out through the corresponding coefficients. The specific expressions depend on the adopted differencing scheme.…”

Section: Governing Equations and Numerical Methodsmentioning

confidence: 99%

“…Over the past five years, the number of non-graphical applications that use GPUs to perform heavy computations has significantly increased in a variety of fields [6,7]. However, as reported recently [4] there are some performance limitations in regard to Computational Fluid Dynamics (CFD).…”

Section: Introductionmentioning

confidence: 99%

“…Brandvik and Pullan [9] reported speed-ups of 29x (2D) and 16x (3D) when solving the Euler equations using structured grids. Elsen et al [7] also solved the Euler equations for hypersonic flows with speed-ups of 40x (simple geometries) or 20x (complex geometries). However, all these performances were attained by using single precision (SP) codes, because the first generation of GPU hardware did not allow double precision (DP) computations.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

On the performance of a 2D unstructured computational rheology code on a GPU

Pereira

Vuik

Pinho

et al. 2013

AIP Conference Proceedings

View full text Add to dashboard Cite

Abstract. The present work explores the massively parallel capabilities of the most advanced architecture of graphics processing units (GPUs) code named "Fermi", on a two-dimensional unstructured cell-centred finite volume code. We use the SIMPLE algorithm to solve the continuity and momentum equations that was fully ported to the GPU. The benefits of this implementation are compared with a serial implementation that traditionally runs on the central processing unit (CPU). The developed codes were assessed with the bench-mark problems of Poiseuille flow, for Newtonian and generalized Newtonian fluids, as well as by the lid-driven cavity and the sudden expansion flows for Newtonian fluids. The parallel (GPU) code accelerated the resolution of those three problems by factors of 19, 10 and 11, respectively, in comparison with the corresponding CPU single core counterpart. The results are a clear indication that GPUs are and will be useful in the field of computational fluid dynamics (CFD) for rheologically simple and complex fluids.

show abstract

Section: Governing Equations and Numerical Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

On the performance of a 2D unstructured computational rheology code on a GPU

Pereira

Vuik

Pinho

et al. 2013

AIP Conference Proceedings

View full text Add to dashboard Cite

show abstract

“…[50][51][52] Nevertheless, to achieve such huge accelerations, the numerical approach must take full advantage of the GPU architecture. The smoothed-particle hydrodynamics (SPH) technique has been established as one of the major concepts for fluid animation in computer graphics.…”

Section: Process Simulations On Gpusmentioning

confidence: 99%

Simulation in Metallurgical Processing: Recent Developments and Future Perspectives

Ludwig

Kharicha

2016

JOM

View full text Add to dashboard Cite

This article briefly addresses the most important topics concerning numerical simulation of metallurgical processes, namely, multiphase issues (particle and bubble motion and flotation/sedimentation of equiaxed crystals during solidification), multiphysics issues (electromagnetic stirring, electro-slag remelting, Cu-electro-refining, fluid-structure interaction, and mushy zone deformation), process simulations on graphical processing units, integrated computational materials engineering, and automatic optimization via simulation. The present state-of-the-art as well as requirements for future developments are presented and briefly discussed.

show abstract

“…Its low cost, high floating-point-operation throughput, and memory access bandwidth have been attracting more and more researchers in the field of high-performance computing [7,[31][32][33][34]. In addition, compared with cluster systems that consist of many CPUs, GPU computing is low-cost and requires low energy at equivalent performance.…”

Section: Graphics Processing Unitsmentioning

confidence: 99%

A High-Order-Accurate GPU-Based Radiative Transfer Equation Solver for Combustion and Propulsion Applications

Lee

Wilcox

et al. 2013

Numerical Heat Transfer, Part B: Fundamentals

View full text Add to dashboard Cite

In this article we present a high-order-accurate solver for the radiative transfer equation (RTE) which uses the discontinuous Galerkin (DG) method and is designed for graphics processing units (GPUs). The compact nature of the high-order DG method enhances scalability, particularly on GPUs. High-order spatial accuracy can be used to reduce discretization errors on a given computational mesh, and can also reduce the mesh size needed to achieve a desired error tolerance. Computational efficiency is a key concern in solutions to radiative heat transfer problems, due to potentially large problem sizes created by (a) the presence of participating nongray media in a full-spectrum analysis, (b) the need to resolve a large number of angular directions and spatial extent of the domain for an accurate solution, and (c) potentially large variations in material and flow properties in the domain. We present here a simulation strategy, as well as a set of physical models, accompanied by a number of case studies, demonstrating the accuracy and superior performance in terms of computational efficiency of this approach.

show abstract

Large calculation of the flow over a hypersonic vehicle using a GPU

Cited by 211 publications

References 18 publications

On the performance of a 2D unstructured computational rheology code on a GPU

On the performance of a 2D unstructured computational rheology code on a GPU

Simulation in Metallurgical Processing: Recent Developments and Future Perspectives

A High-Order-Accurate GPU-Based Radiative Transfer Equation Solver for Combustion and Propulsion Applications

Contact Info

Product

Resources

About