Abstract
Elastic Domain Decomposition Method (DDM) for a parallel Adaptive Mesh Refinement (AMR) Solver is described. To use this method saving a message passing cost, a good load balancing, and a high parallel performances (73% on 32PEs) were achieved.
