A progressive mesh method for physical simulations using lattice Boltzmann method on single-node multi-gpu architectures

TL;DR

This paper introduces a progressive mesh algorithm leveraging multi-GPU architectures to perform fast, adaptive fluid simulations with the lattice Boltzmann method on complex geometries, outperforming static mesh approaches.

Contribution

The paper presents a novel progressive mesh algorithm that adapts mesh resolution dynamically for efficient fluid simulations on single-node multi-GPU systems.

Findings

Achieves significant performance improvements over static mesh methods.

Successfully simulates complex geometries with adaptive meshing.

Demonstrates scalability on multi-GPU architectures.

Abstract

In this paper, a new progressive mesh algorithm is introduced in order to perform fast physical simulations by the use of a lattice Boltzmann method (LBM) on a single-node multi-GPU architecture. This algorithm is able to mesh automatically the simulation domain according to the propagation of fluids. This method can also be useful in order to perform various types of simulations on complex geometries. The use of this algorithm combined with the massive parallelism of GPUs allows to obtain very good performance in comparison with the static mesh method used in literature. Several simulations are shown in order to evaluate the algorithm.