itaps-parallel IGeomP, iRelP?

[Tim Tautges]

There are various methods for distributing the model, and in our current 
API these are already accessible (using the option string to the load 
function).  Contrary to iMeshP, I don't see much of a need to do 
parallel communication of model entities once the load is done.  So, I 
doubt we need a parallel interface for geometry.  Same goes for 
relations.  We definitely will need one for fields, though.

- tim

Mark Shephard wrote:
> The CAD models are typically small enough that we have not worried about 
> distributing them also. If we need to consider distributing the model 
> there are a number of possible approaches with the one that I would push 
> being a function of what applications were being supported. For example 
> I would tend to use different methods for distributing the model in each 
> of the three cases of doing modeling operations, supporting mesh 
> generation from the model and supporting mesh adaptation/analysis.
> 
> If we think its critical, it is something I can discuss.
> 
> 
> Devine, Karen D wrote:
>> Mark and Tim (and anyone else who wants to comment):
>>
>> In our original ITAPS proposal, we mention partitioning of geometry
>> information.  We have focused on parallel meshes so far.  But what are 
>> your
>> needs for distributed geometry information and mapping between 
>> distributed
>> geometries and meshes?  How do you handle this information now?  What 
>> do you
>> envision you and your applications people will need going forward?
>>
>> Thanks.
>> Karen
>>
>>
> 
> 

-- 
================================================================
"You will keep in perfect peace him whose mind is
   steadfast, because he trusts in you."               Isaiah 26:3

              Tim Tautges            Argonne National Laboratory
          (tautges at mcs.anl.gov)      (telecommuting from UW-Madison)
          phone: (608) 263-8485      1500 Engineering Dr.
            fax: (608) 263-4499      Madison, WI 53706