Parallel multigrid technique: reduction to independent problems
Parallel multigrid computation algorithms have nowadays more and more wide applications in spite of the fact that they show some limitations. First of all there is a nonbalanced loading of the processors as well as the intensive data exchange between processors. A version of modification of the universal parallel multigrid computation technique is proposed. The modification consists of a reduction of the initial difference boundary problem to a set of independent problems. A finite difference scheme of the six order of approximation is taken for test experiments.
- Structure of highly parallel, efficient, scalable, true robust pseudomultigrid technique for black-box solving a large class of the boundary value problems on high performance computing systems
- Potentialities of the robust multigrid technique
- Segmental refinement: a multigrid technique for data locality
- A parallel domain reduction method
This page was built for publication: Parallel multigrid technique: reduction to independent problems
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2830695)