Out-of-core parallel Delaunay mesh generation

Abstract We present an extension of an existing Parallel Delaunay Re¯nement (PDR) method to be used for out-of-core computing, thus allowing it to generate large meshes using limited computing resources. We base our work on a shared memory implementation of the PDR method. Also, we are designing a distributed out-of-core algorithm which would allow to compute very large meshes using supercom-puters and COWs with limited resources