Subject: Re: [Libmesh-users] pre-partitioning a big mesh

Hi John,

This is the first time I see this. So I’m not sure what this .cpr format is?
Can each processor read its own chunk using with when libMesh is
compiled with parallel mesh?


> On Jan 11, 2017, at 3:32 PM, John Peterson <[email protected]> wrote:
> On Wed, Jan 11, 2017 at 1:29 PM, gmail <[email protected]
> <mailto:[email protected]>> wrote:
> Hi all,
> I am working with a relatively large 3D exodusII meshes 15-50 M elements that
> I create using Trelis (CUBIT). Right now I have crude procedure to
> pre-partition this mesh on a machine with a large memory basically by reading
> it in a small libMesh program and then writing it out as a nemesis series.
> This is not ideal and I would like to find a way to pre-partition this mesh
> on my office machine (or a small cluster) for any number of CPUs.
> I searched through the mailing list and have found this thread:
> <>
> <
> <>> that suggests
> use of SEACAS or using meshtools to save the mesh into a...

n XDR format. I had

> little success with the first method since the documentation for SEACAS are
> incomplete. I managed to create a partition using nem_slice but could not
> make nem_spread work (probably not doing the inp file correctly).
> I would like to ask the developers as well as the members of mailing list to
> comment and share their procedure dealing with large meshes.
> Have you had a chance to try the src/apps/splitter.C code. It was recently
> added by Derek to do exactly what you are asking about.
> --
> John

Developer Access Program for Intel Xeon Phi Processors
Access to Intel Xeon Phi processor-based developer platforms.
With one year of Intel Parallel Studio XE.
Training and support from Colfax.
Order your platform today.
Libmesh-users mailing list
[email protected]

Programming list archiving by: Enterprise Git Hosting