Archived issue #0006729
Slow performance of ShapeAnalysis_FreeBounds
Description
The computation time of ShapeAnalysis_FreeBounds in a great degree depends on
the number of wires generated by this algorithm. Attached, please, find the
model and draw script representing the problem.
It gives the following figures on SUN Ultra 10 in debug mode:
....
nb of wires = 11
CPU user time: 0.83 seconds
nb of wires = 80
CPU user time: 33.47 seconds
nb of wires = 263
CPU user time: 80.83 seconds
....
It is needed to analyze the problem and to try to improve performance of the
algorithm.
the number of wires generated by this algorithm. Attached, please, find the
model and draw script representing the problem.
It gives the following figures on SUN Ultra 10 in debug mode:
....
nb of wires = 11
CPU user time: 0.83 seconds
nb of wires = 80
CPU user time: 33.47 seconds
nb of wires = 263
CPU user time: 80.83 seconds
....
It is needed to analyze the problem and to try to improve performance of the
algorithm.
Additional information
Documentation remark, added by AMV 2005-03-03 08:25:28:
Improvements:
The algorithm ShapeAnalysis_FreeBounds has been optimized for performance by
means of avoidance of quadratic dependence of the computation time on the number
of free
boundaries.
The algorithm BRepAlgo_Sewing has been improved so as to generate the same
stable result repeatedly on the same input data.
Improvements:
The algorithm ShapeAnalysis_FreeBounds has been optimized for performance by
means of avoidance of quadratic dependence of the computation time on the number
of free
boundaries.
The algorithm BRepAlgo_Sewing has been improved so as to generate the same
stable result repeatedly on the same input data.
Public activity
No public notes
Participants are labeled by their role within this record.