<div dir="ltr"><div class="gmail_extra"><div class="gmail_quote">On Wed, May 11, 2016 at 2:45 AM, Åsmund Ervik <span dir="ltr"><<a href="mailto:asmund.ervik@ntnu.no" target="_blank">asmund.ervik@ntnu.no</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">Dear all (Tobin and Matt in particular),<br>
<br>
I have some questions about DMForest, sparked by a mention of this in an<br>
earlier email to the list by Matt.<br>
<br>
As I understand it, DMForest is a DM which aims to provide a<br>
hierarchically refined grid, i.e. structured adaptive mesh refinement<br>
(AMR). Correct me if I'm wrong here.<br>
<br>
Now, my questions:<br></blockquote><div><br></div><div>Toby is more qualified to answer, but I will start:</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
What is the level of maturity of DMForest? Could I start using it today,<br>
or would it be better to wait for upcoming features/fixes?<br></blockquote><div><br></div><div>You can use DMForest purely as a Plex now, and I would characterize this</div><div>as ready to be tried out. If you want to use it mainly as a forest (structured</div><div>adaptive mesh), then I would say its alpha.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
To what extent is DMForest intended to be a drop-in replacement for<br>
DMDA? For use in a CFD code already using DMDA, let's say finite<br>
difference, explicit time-integration, compressible Navier-Stokes should<br>
I expect switching to be a major undertaking? How encapsulated would the<br>
changes be?<br></blockquote><div><br></div><div>Not at all. The major advantage of DMDA is the regular topology (gone) and</div><div>the nice array interface (not there either). DMForest is mainly focused on</div><div>cell-focused discretizations like DG, and also more traditional FEM. It is no</div><div>coincidence that this is also how people use p4est on its own.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
Is DMForest intended to provide dynamic adaptivity? Let's say I'm<br>
considering a fluid jet which creates a Kelvin-Helmholtz instability,<br>
and I want to refine the mesh to resolve the vortices appearing. Is this<br>
feasible? If possible, I assume there is some limitation on how<br>
frequently one can re-adapt to maintain good performance?<br></blockquote><div><br></div><div>Yes. I would say its intended to do very frequent adaptation.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
What are my options for load-balancing with DMForest? Suppose I have<br>
some process coupled with the flow (e.g. chemistry, thermodynamics) that<br>
takes a significant amount of computing time, but not in a uniform way<br>
across all grid cells, is there (or could there be) an interface for<br>
specifying "these cells will have to do expensive calculations, these<br>
cells will do less expensive calculations"?<br></blockquote><div><br></div><div>Both Forest and Plex will accept weights for partitioning. Its not quite clear what</div><div>the final interface should be, but we would help you to get your weights in.</div><div><br></div><div> Thanks,</div><div><br></div><div> Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
This latter point would IMO also be useful for DMDA/DMPlex, so maybe it<br>
could be made generic? The biggest problem I guess is the tradeoff<br>
between the interface being simple enough and powerful enough. Perhaps a<br>
"good enough" option would be some sort of autotuning based on logging<br>
the time taken to arrive at an MPI sync point for all the MPI processes,<br>
and offloading work from the slowest ones to their neighbours.<br>
<br>
<br>
Thanks,<br>
Åsmund<br>
<br>
</blockquote></div><br><br clear="all"><div><br></div>-- <br><div class="gmail_signature">What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>-- Norbert Wiener</div>
</div></div>