Is there calc-speed-difference due to OCCT-ver.-differences?

Takeshi Fujita
Mon, 02/20/2017 - 09:57

Forums:

Other usage issues

I calculate intersection point of two objects by BRepExtrema_DistShapeShape and PointOnShape().

Calc-speed via OCCT Ver6.6.0-x86 is three times slower than that via OCCT ver7.0.0-X64.

Is this calc-speed-difference due to OCCT-ver.-differences or due to 32vs64bit-differences?

Thanks in advance ;D

Benjamin Bihler
Mon, 02/20/2017 - 11:51

According to my experiences both! But the improvements of the newer OCCT versions have the higher impact.

Takeshi Fujita
Tue, 02/21/2017 - 02:33

Hi, Benjamin Bihler :D

Thank you for your comment !
Your comment seems to be exactly correct.
I’m going to use OCCT newer version instead of using older version.

Forum supervisor
Tue, 02/21/2017 - 11:19

Dear Takeshi,

64-bit builds usually show better performance over 32-bit builds, but not in times (as far as no other building options differ).
The changes you see are most likely due to optimizations in extrema algorithms within newer versions of Open CASCADE Technology.
You can read a couple of articles related to these optimizations on Development Portal:

https://dev.opencascade.org/index.php?q=node/988
https://dev.opencascade.org/index.php?q=node/975

Best Regards, FSR

Takeshi Fujita
Tue, 02/21/2017 - 11:45

Dear Forum supervisor,

Thank you for additional detailed information.

These articles are very informative.

Thank you again,