Two-level parallelization of a fluid mechanics algorithm exploiting hardware heterogeneity

DOI10.1016/j.compfluid.2015.05.012zbMath1390.76008OpenAlexW1038848389MaRDI QIDQ1645828

Jörg Stiller, Immo Huismann, Jochen Fröhlich

Publication date: 22 June 2018

Published in: Computers and Fluids (Search for Journal in Brave)

Full work available at URL: https://doi.org/10.1016/j.compfluid.2015.05.012

zbMATH Keywords

MPIparallelizationOpenMPheterogeneous computingGPGPUOpenACC

Mathematics Subject Classification ID

Spectral methods applied to problems in fluid mechanics (76M22) Parallel numerical computation (65Y05) Spectral, collocation and related methods for initial value and initial-boundary value problems involving PDEs (65M70) Software, source code, etc. for problems pertaining to fluid mechanics (76-04)

Related Items (3)

Multilevel parallelization for simulating compressible turbulent flows on most kinds of hybrid supercomputers ⋮ Scaling to the stars -- a linearly scaling elliptic solver for \(p\)-multigrid ⋮ A THREE-LEVEL PARALLELISATION SCHEME AND APPLICATION TO THE NELDER-MEAD ALGORITHM

Uses Software

• CUDA
• PETSc
• MPI
• GPGPU
• OpenACC
• StarPU
• SkePU
• HOSTA
• OmpSs

Cites Work

• Unnamed Item
• Collaborating CPU and GPU for large-scale high-order CFD simulations with complex grids on the TianHe-1A supercomputer
• An improved immersed boundary method with direct forcing for the simulation of particle laden flows
• Nodal discontinuous Galerkin methods on graphics processors
• A spectral element method for fluid dynamics: Laminar flow in a channel expansion
• Parallel computing with load balancing on heterogeneous distributed systems.
• Optimizing 2D and 3D structured Euler CFD solvers on graphical processing units
• High-order finite-element seismic wave propagation modeling with MPI on a large GPU cluster
• Partially Implicit BDF2 Blends for Convection Dominated Flows
• High-Order Methods for Incompressible Fluid Flow
• From h to p efficiently: optimal implementation strategies for explicit time‐dependent problems using the spectral/hp element method

This page was built for publication: Two-level parallelization of a fluid mechanics algorithm exploiting hardware heterogeneity