Portable implementation model for CFD simulations. Application to hybrid CPU/GPU supercomputers

DOI10.1080/10618562.2017.1390084OpenAlexW2765368849MaRDI QIDQ5072891

Assensi Oliva, A. V. Gorobets, Ricard Borrell, Guillermo Oyarzun

Publication date: 5 May 2022

Published in: International Journal of Computational Fluid Dynamics (Search for Journal in Brave)

Full work available at URL: http://hdl.handle.net/2117/121221

zbMATH Keywords

OpenCLheterogeneous computingGPUportabilityCFD codeMPI+CUDAsliced ELLPACK

Mathematics Subject Classification ID

Fluid mechanics (76-XX) Numerical analysis (65-XX)

Related Items

HPC\(^2\) -- a fully-portable, algebra-based framework for heterogeneous computing. Application to CFD, On the implementation of flux limiters in algebraic frameworks, A scalable framework for the partitioned solution of fluid-structure interaction problems, A hierarchical parallel implementation for heterogeneous computing. Application to algebra-based CFD simulations on hybrid supercomputers, A GPU-based algorithm for efficient LES of high Reynolds number flows in heterogeneous CPU/GPU supercomputers

Uses Software

• CUDA
• CUBLAS
• CUSPARSE
• TERMOFLUIDS

Cites Work

• Symmetry-preserving discretization of Navier-Stokes equations on collocated unstructured grids
• CFD-based analysis and two-level aerodynamic optimization on graphics processing units
• MPI-CUDA sparse matrix-vector multiplication for the conjugate gradient method with an approximate inverse preconditioner
• Symmetry-preserving discretization of turbulent flow.
• Subgrid-scale stress modelling based on the square of the velocity gradient tensor
• Large calculation of the flow over a hypersonic vehicle using a GPU
• Solving incompressible two-phase flows on multi-GPU clusters
• A portable OpenCL-based unstructured edge-based finite element Navier-Stokes solver on graphics hardware
• Direct numerical simulation of the flow over a sphere at Re = 3700
• Performance Optimization Using Partitioned SpMV on GPUs and Multicore CPUs
• Unsteady CFD computations using vertex-centered finite volumes for unstructured grids on Graphics Processing Units
• Optimising the Termofluids CFD code for petascale simulations
• Numerical Solution of the Navier-Stokes Equations