An efficient GPU-accelerated adaptive mesh refinement framework for high-fidelity compressible reactive flows modeling

TL;DR

This paper introduces a GPU-accelerated adaptive mesh refinement framework for high-fidelity reactive flow simulations, achieving significant speedups and scalability while maintaining accuracy in complex physics problems.

Contribution

The paper develops a novel GPU-accelerated AMR framework with specialized algorithms and low-storage chemical integrators, enabling efficient large-scale reactive flow simulations.

Findings

Achieved up to 6.49x speedup on GPU compared to CPU.

Demonstrated excellent parallel scalability across multiple GPU nodes.

Successfully simulated large-scale reactive shock-bubble interactions with reduced computational costs.

Abstract

This paper presents a heterogeneous adaptive mesh refinement (AMR) framework for efficient simulation of moderately stiff reactive problems. This framework features an elaborate subcycling-in-time algorithm along with a specialized refluxing algorithm, all unified in a highly parallel codebase. We have also developed a low-storage variant of explicit chemical integrators by optimizing the register usage of GPU, achieving respectively 6x and 3x times speedups as compared to the implicit and standard explicit methods with comparable order of accuracy. A suite of benchmarks have confirmed the framework's fidelity for both non-reactive and reactive simulations with/without AMR. By leveraging our parallelization strategy that is developed on AMReX, we have demonstrated remarkable speedups on various problems on a NVIDIA V100 GPU than using a Intel i9 CPU within the same codebase; in problems with complex physics and spatiotemporally distributed stiffness such as the hydrogen detonation propagation, we have achieved an overall 6.49x acceleration ratio. The computation scalability of the framework is also validated through the weak scaling test, demonstrating excellent parallel efficiency across multiple GPU nodes. At last, a practical application of this GPU-accelerated SAMR framework to large-scale direct numerical simulations is demonstrated by successful simulation of the three-dimensional reactive shock-bubble interaction problem; we have saved significant computational costs while maintaining the comparable accuracy, as compared to a prior uniform DNS study performed on CPUs.