Adaptive kinetic-fluid solvers for heterogeneous computing architectures
From MaRDI portal
Publication:2374880
DOI10.1016/J.JCP.2015.10.003zbMATH Open1349.76754arXiv1503.00707OpenAlexW1934316561MaRDI QIDQ2374880FDOQ2374880
Authors: Sergey Zabelok, Robert Arslanbekov, Vladimir Kolobov 
Publication date: 5 December 2016
Published in: Journal of Computational Physics (Search for Journal in Brave)
Abstract: We show feasibility and benefits of porting an adaptive multi-scale kinetic-fluid code to CPU-GPU systems. Challenges are due to the irregular data access for adaptive Cartesian mesh, vast difference of computational cost between kinetic and fluid cells, and desire to evenly load all CPUs and GPUs. Our Unified Flow Solver (UFS) combines Adaptive Mesh Refinement (AMR) with automatic cell-by-cell selection of kinetic or fluid solvers based on continuum breakdown criteria. Using GPUs enables hybrid simulations of mixed rarefied-continuum flows with a million of Boltzmann cells with 24x24x24 velocity mesh. We describe the implementation of CUDA kernels for three modules in UFS: the direct Boltzmann solver using discrete velocity method, the Direct Simulation Monte Carlo (DSMC) solver, and a mesoscopic solver based on Lattice Boltzmann Method, all using octree Cartesian mesh. Double digit speedups on single GPU and good scaling for multi-GPUs have been demonstrated.
Full work available at URL: https://arxiv.org/abs/1503.00707
Recommendations
- Solving model kinetic equations on GPUs
- Multi-block adaptive mesh refinement (AMR) for a lattice Boltzmann solver using gpus
- Towards an ultra efficient kinetic scheme. part III: high-performance-computing
- Solving the Boltzmann equation on GPUs
- GPU accelerated cell-based adaptive mesh refinement on unstructured quadrilateral grid
MPIlattice Boltzmann methodadaptive mesh refinementBoltzmann kinetic equationCUDAgraphics processing unitsdiscrete velocity methoddirect simulation Monte Carlounified flow solver
Parallel numerical computation (65Y05) Probabilistic methods, particle methods, etc. for initial value and initial-boundary value problems involving PDEs (65M75) Particle methods and lattice-gas methods (76M28) Numerical algorithms for specific classes of architectures (65Y10)
Cites Work
- LUDWIG: A parallel lattice-Boltzmann code for complex fluids
- Hemelb: A high performance parallel lattice-Boltzmann code for large scale fluid flow in complex geometries
- Sailfish: a flexible multi-GPU implementation of the lattice Boltzmann method
- Numerical methods for kinetic equations
- Unified solver for rarefied and continuum flows with adaptive mesh and algorithm refinement
- Kinetic theory representation of hydrodynamics: a way beyond the Navier–Stokes equation
- Direct methods for solving the Boltzmann equation and study of nonequilibrium flows
- A survey on parallel computing and its applications in data-parallel problems using GPU architectures
- A Cartesian cut cell method for rarefied flow simulations around moving obstacles
- Adaptive mesh and algorithm refinement using direct simulation Monte Carlo
- Adaptive kinetic-fluid solvers for heterogeneous computing architectures
- Numerical simulation of a moving rigid body in a rarefied gas
- Solving model kinetic equations on GPUs
- A lattice-Boltzmann method with hierarchically refined meshes
- Towards adaptive kinetic-fluid simulations of weakly ionized plasmas
- Coupled Vlasov and two-fluid codes on GPUs
- Finite volume TVD formulation of lattice Boltzmann simulation on unstructured mesh
- A multiscale kinetic-fluid solver with dynamic localization of kinetic effects
- Parallel simulation of multiphase flows using octree adaptivity and the volume-of-fluid method
- On efficient simulations of multiscale kinetic transport
- A GPU-CUDA based direct simulation Monte Carlo algorithm for real gas flows
Cited In (16)
- A Cartesian cut cell method for rarefied flow simulations around moving obstacles
- Advanced parallelization strategies using hybrid MPI-CUDA octree DSMC method for modeling flow through porous media
- Adaptive kinetic-fluid solvers for heterogeneous computing architectures
- A weighted multiple-relaxation-time lattice Boltzmann method for multiphase flows and its application to partial coalescence cascades
- Local Discrete Velocity Grids for Multi-Species Rarefied Flow Simulations
- Application of adaptively refined unstructured grids in DSMC to shock wave simulations
- Level-by-level artificial viscosity and visualization for MHD simulation with adaptive mesh refinement
- GPU acceleration of an iterative scheme for gas-kinetic model equations with memory reduction techniques
- Hybrid lattice Boltzmann-direct simulation Monte Carlo approach for flows in three-dimensional geometries
- A compressed lattice Boltzmann method based on ConvLSTM and resnet
- Numerical simulation of a moving rigid body in a rarefied gas
- Solving model kinetic equations on GPUs
- A new cut-cell algorithm for DSMC simulations of rarefied gas flows around immersed moving objects
- Discontinuous Galerkin algorithms for fully kinetic plasmas
- Computationally efficient high-fidelity plasma simulations by coupling multi-species kinetic and multi-fluid models on decomposed domains
- An efficient numerical method for solving the Boltzmann equation in multidimensions
Uses Software
This page was built for publication: Adaptive kinetic-fluid solvers for heterogeneous computing architectures
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2374880)
