Implement overlapping sub-grids optimisations

[MatveyK]

Tha system currently struggles with large outputs due to undefined states and possibly slow computation time. It thus could be good idea to divide the large output into small sub-grids and resolve them separately from one another. In order to preserve global coherence we would just need to overlap the sub-grids so that they can propagate the information between them.

It could also provide an easier framework for eventual multithreading work.