High Performance Computing in Fluid Dynamics

This book contains the course notes of the Summerschool on High Per formance Com;~ting in Fluid Dynamics, held at the Delft University of Technology, June 24-28, 1996, under the auspices of and with support of ERCOFTAC (European Research Community on Flow, Turbulence and Combustion, WWW: and the J.M. Burgers Center (Graduate School for Fluid Dynamics, The Nether lands, WWW: In addition to the material presented in the present volume, lectures were given by J. Hauser on "A parallel Newton-GMRES method for the 3D N avier-Stokes equations in complex geometries". The material written down by F.-S. Lien was presented by M.A. Leschziner. This book is addressed to students on a graduate level and researchers in industry engaged in scientific computing, who have little or no experience with high performance computing, but who want to learn more, and/or want to port their code to parallel platforms. Applications in computa tional fluid dynamics are emphasized.

Texte du rabat

This book contains the course notes of the Summerschool on High Performance Computing in Fluid Dynamics, held at the Delft University of Technology, June 24-28, 1996. The lectures presented deal to a large extent with algorithmic, programming and implementation issues, as well as experiences gained so far on parallel platforms. Attention is also given to mathematics aspects, notably domain decomposition and scalable algorithms. Topics considered are: basic concepts of parallel computers, parallelization strategies, programming aspects, parallel algorithms, applications in computational fluid dynamics, the present hardware situation and developments to be expected. The book is addressed to students on a graduate level and researchers in industry engaged in scientific computing, who have little or no experience with high performance computing, but who want to learn more, and/or want to port their code to parallel platforms. It is a good starting point for those who want to enter the field of high performance computing, especially if applications in fluid dynamics are envisaged.

Contenu

1. Introduction to Parallel Computers: Architecture and Algorithms.- 2. Programming Aspects and Algorithms for Vector- and Parallel Computers.- 3. Porting Industrial Codes on High-Performance Computers.- 4. Load Balancing for Computational Fluid Dynamics.- 5. Parallel Linear Systems Solvers: Sparse Iterative Methods.- 6. High Performance Computing of Turbulent Flows.- 7. Data-Parallel Solution of the Incompressible Navier-Stokes Equations.- 8. High Performance Computing: Trends and Expectations.