[cgma-dev] Large OCC subtraction time

[Jiangtao Hu]

Hi, Rajeev:
 
Thank you for providing the testcase. OCC did have a reputation of getting much slower with larger models and more operations. I'll have a look on Monday to see if there's anything I can do on cgm side that can speed it up.
 
Have a nice weekend!
 
Jane

--- On Fri, 2/25/11, Rajeev Jain <jain at mcs.anl.gov> wrote:


From: Rajeev Jain <jain at mcs.anl.gov>
Subject: [cgma-dev] Large OCC subtraction time
To: cgma-dev at mcs.anl.gov
Date: Friday, February 25, 2011, 4:21 PM





When subtracting 325 pins from a block, OCC takes a lot more time than the corresponding ACIS model. I'm attaching the assygen .inp that can be run as MeshKit/rgg>assygen 7.  This file is for OCC version of MeshKit. For ACIS change the geometry engine to ACIS in the .inp file and run ACIS version.

RGG o/p:

ACIS:

Normal Termination.
Geometry file: 7.sat saved.
Schemes file created: 7.template.jou
Cubit journal file created: 7.jou
Ending at : Fri Feb 25 15:10:31 2011
Elapsed wall clock time: 8 seconds or 0.133333 mins


OCC: 

Exported: 327 OCC Bodies to 7.stp

 Normal Termination.
Geometry file: 7.stp saved.
Schemes file created: 7.template.jou
Cubit journal file created: 7.jou
Ending at : Fri Feb 25 14:59:51 2011
Elapsed wall clock time: 3173 seconds or 52.8833 mins

The geometry files can be found at /homes/jain/MeshKit-all/trunk/rgg/test/geometry_parallel_meshing/7.stp and 7.sat

Rajeev






      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/cgma-dev/attachments/20110226/f9216271/attachment.htm>