On Wed, 23 Mar 2011, Roy Stogner wrote: > On Fri, 18 Mar 2011, fdkong wrote: > >> I am developing a program to solve physical problems. I want to >> use parallel mesh of libmesh, but I do not know whether parallel >> mesh support large-scale cluster; that is, can I use 100000 cpus >> with using prarallel mesh? > > In theory, yes. In practice, ParallelMesh currently has bugs with > adaptive coarsening, which means that none of the main developers > regularly use it, which means that it's likely to have other bugs that > just haven't been found yet.
I should also point out that this "current" status is likely to change over the next year - UT is starting to worry about exascale computing, and only part of that scaling is going to be accomplished at a higher uncertainty quantification level; we'd like the rest to come from getting ParallelMesh up to speed. INL is in the same boat, I believe. As a first step, I just fixed some of the serializing I/O classes that had regressed over the last year; "make run_examples" with the --enable-parmesh configuration should be working again, and it's now in continuous integration so we'll notice if it gets re-broken. --- Roy ------------------------------------------------------------------------------ Create and publish websites with WebMatrix Use the most popular FREE web apps or write code yourself; WebMatrix provides all the features you need to develop and publish your website. http://p.sf.net/sfu/ms-webmatrix-sf _______________________________________________ Libmesh-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/libmesh-users
