Actually, there is _some_ support for distributed IO on exodusII files using 
DMMesh. I assume that porting the existing code to DMComplex would not be too 
difficult, but I don't have time to work on it right now.
The mesh geometry initial reading is sequential, but distributed meshes and 
result files can be written. The caveat is that there is no interface matching, 
which can be a problem with some post processing functions.

Blaise

Sent from a mobile device

On Nov 19, 2012, at 5:33 PM, "Matthew Knepley" <knepley at gmail.com> wrote:

> On Mon, Nov 19, 2012 at 6:27 PM, Fande Kong <fande.kong at colorado.edu> 
> wrote:
>> Hi all,
>> 
>> Are there functions that can be used to implement parallel read and write?
> 
> Now, all the reads are in serial right now. Is this a bottleneck?
> 
>   Matt
> 
>> --
>> Fande Kong
>> Department of Computer Science
>> University of Colorado at Boulder
> 
> 
> 
> --
> What most experimenters take for granted before they begin their
> experiments is infinitely more interesting than any results to which
> their experiments lead.
> -- Norbert Wiener
> 

Reply via email to