Application of parallel programming tools for the solution of fluid dynamics problems on multi-processor computing systems

Some questions connected with the solution of fluid dynamics problems on multi-processor computing systems are considered. The methods of domain decomposition, the ways of data distribution on processors, and special features of parallel implementation of the numerical methods used for the solution of particular problems are discussed. Several examples of the solution of some problems of fluid and gas dynamics on structured and unstructured grids are discussed and the performance of program codes is studied depending on the size of a problem and on the number of processors.