I have just read the page and I have one more concern to make things clear. Will dealII need to read in the whole mesh in one node and do the partition work and then redistribute the mesh partitions to different nodes? I guess this will influence the scale of the problem can be solved.
It will need to read the entire coarse mesh on *every* node, but you can then refine it in a distributed way. That obviously limits the size of the coarse mesh (we've done tests with a couple of 100,000 coarse mesh cells) but not of the fine mesh (we've tested to 10^9 or so cells).
Best W. ------------------------------------------------------------------------ Wolfgang Bangerth email: [email protected] www: http://www.math.tamu.edu/~bangerth/ _______________________________________________ dealii mailing list http://poisson.dealii.org/mailman/listinfo/dealii
