← Back to team overview

yade-users team mailing list archive

Re: [Question #704367]: OpenMp

 

Question #704367 on Yade changed:
https://answers.launchpad.net/yade/+question/704367

Jérôme Duriez posted a new comment:
You may note that the increase in speed with respect to the chosen "n"
is clearly not monotonous, generally speaking. And that the optimum
could be well below 24 (or well above..) depending on many factors such
as the type of YADE simulation at hand.

You have examples in
https://www.sciencedirect.com/science/article/abs/pii/S0266352X21000367
§ 5 (Sphere or more complex shapes) or
https://www.sciencedirect.com/science/article/pii/S0098300421002247 §
6.3 (complex shapes only). I will not claim generality for the observed
behaviors therein but it may give some ideas.

-- 
You received this question notification because your team yade-users is
an answer contact for Yade.