On 6/20/19 9:19 AM, Stephen DeWitt wrote:
> I'm running matrix free calculations using a
> parallel::distributed::Triangulation (loosely based on step-48). One thing
> that I've noticed is the the final number of DOFs at the end of a simulation
> vary slightly depending on my number of MPI processes. For example, one run
> on
> 16 cores finished with 2,988,700 DOF and nominally the same run on 512 cores
> finished with 2,988,864. Is this to be expected?
If the number of cells is also different, then yes. For the same number of
cells, you need to get the same number of DoFs, so the number of cells is the
normative question you need to investigate.
If you do use adaptive mesh refinement, then just as you mention there are
heuristics that sometimes do or do not lead to certain cells to be refined.
For example, computing the threshold for the local error indicators above
which a cell should be refined is one algorithm that might slightly depend on
the number of processes you use.
Best
W.
--
------------------------------------------------------------------------
Wolfgang Bangerth email: bange...@colostate.edu
www: http://www.math.colostate.edu/~bangerth/
--
The deal.II project is located at http://www.dealii.org/
For mailing list/forum options, see
https://groups.google.com/d/forum/dealii?hl=en
---
You received this message because you are subscribed to the Google Groups
"deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to dealii+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/dealii/e2f7f0ac-a2ba-04b6-4ae3-a5fcd7342bed%40colostate.edu.
For more options, visit https://groups.google.com/d/optout.
