Skip to content
Preprint

A GPU-efficient local ADER finite-difference method for compressible-flow DNS

Oct 2026 · 0 citations · 37 references
Physics

TL;DR

A local arbitrary high-order derivatives (ADER) finite-difference method that couples a compact ten-point quadratic predictor to a conservative central corrector through operator matching recovers third-order convergence under fixed-CFL refinement and near-seventh-order spatial behavior when background spatial error dominates.

Abstract

Algorithm design tailored to graphics processing units (GPUs) can substantially improve the efficiency of computational fluid dynamics (CFD) simulations. Efforts to adapt CFD solvers to GPUs often focus on spatial discretization, while time advancement commonly relies on multistage Runge--Kutta (RK) methods. However, these RK methods require repeated residual evaluations and full-domain operations, increasing the cost of compressible-flow direct numerical simulation (DNS). To reduce this overhead, we develop a local arbitrary high-order derivatives (ADER) finite-difference method that couples a compact ten-point quadratic predictor to a conservative central corrector through operator matching. With two quadrature nodes, one Picard correction, and selective dissipation, smooth convective tests recover third-order convergence under fixed-CFL refinement and near-seventh-order spatial behavior when background spatial error dominates. Temporal states remain kernel-local, and halo communication overlaps interior computation. Tests of smooth advection, periodic turbulence, and wall-bounded flows assess accuracy and flow statistics. On a $302^3$-node viscous Taylor--Green benchmark without heat conduction, throughput reaches 230.87 million node updates per second on one V100 GPU in double precision, 1.944 times that of spatially matched three-stage Runge--Kutta integration. Weak-scaling efficiency reaches 99.44\% from 4 to 32 V100 GPUs at 1.812 billion nodes.

View source

Similar papers

Sep 2026

Mixed-precision GPU algorithms for efficient turbulent flow simulations with Raviart-Thomas finite elements

A solution framework with fast matrix-free operator evaluation for all ingredients, combined with multigrid solvers for the Poisson problem, and a robust mixed-precision algorithmic framework to generate accurate initial guesses for the iterative linear solvers are developed.

Ivan Prusak, Enes Mustafa Soydan, Ivan Pribec et al. · 1 citation
Open access Oct 2026

GPU-Parallelization of the Numerical Solution of Large-Scale Transient Partial Differential Equations

Efficient and scalable numerical methods for time-dependent partial differential equations are of central importance. This study extends our previous examinations of OpenCL-based parallel implementation of three explicit finite difference schemes: the one-stage Constant Neighbor (CNe) method, its two-stage predictor–co...

Dániel Koics, E. Kovács, O. Hornyák · 0 citations
Preprint Oct 2026

A Scalable Sharp-Interface Immersed Boundary Method for Moving Boundary Flow Simulation on Multi-GPU Systems

Sharp-interface immersed boundary methods provide a versatile framework for simulating flows with complex moving boundaries, but their irregular data structures, dynamic sparsity, and discontinuous pressure operators present significant challenges for efficient multi-GPU implementation. This work presents a fully GPU-n...

Sushrut Kumar, J. Romero, M. Fatica et al. · 0 citations
Sep 2026

Implicit-Explicit time integration scheme with Physics-based preconditioning for two-fluid tokamak boundary simulations

In this work, a globally stiffly accurate Implicit-Explicit (IMEX) Runge-Kutta scheme is developed and implemented in the GBS code [Ricci et al., Plasma Phys. Control. Fusion, 2012], for two-fluid plasma turbulence simulations. The stiffest phenomena, governed by shear Alfv\'en waves and parallel diffusion, are treated...

Micol Bassanini, Simone Deparis, Paolo Ricci et al. · 0 citations
Open access

Marching faster in time: advanced time-integration for incompressible flows

(English) The computational cost of scale-resolving simulations of incompressible flows, such as Direct Numerical Simulation (DNS) and Large-Eddy Simulation (LES), is largely dictated by the temporal integration of the incompressible Navier-Stokes equations. Traditional timestepping strategies often rely on conservativ...

J. Plana-Riu · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.