MPI Parallelization

[Arpit-Babbar]

This issue discusses high-level goals towards MPI parallelization of TrixiLW.jl. Two mesh types are to be parallelized - TreeMesh and P4estMesh. There are various steps toward getting a production-ready code. These are the goals for the TreeMesh type which is a type of mesh that can only hand rectangular grids -

• MPI Parallelization on uniformly refined TreeMesh with periodic boundary conditions #34
• TreeMesh code with non-periodic boundary conditions (also resolved with the above issue)
• TreeMesh code with adaptive mesh refinement (mortar element method)

The same steps will be there for P4estMesh which is the most powerful mesh type in Trixi.jl

• Parallelizing P4est code on uniformly refined meshes with periodic boundary conditions (Parallel version for dg_2d.jl P4est mesh #37)
• P4est code with non-periodic boundary conditions (also resolved with the above issue)
• P4est code with adaptive mesh refinement (mortar element method)