Regarding the run time of my code

30 views
Skip to first unread message

Md Mahmudul Islam

unread,
Apr 15, 2025, 3:17:02 AM4/15/25
to dea...@googlegroups.com
Hello Everyone,
Hope everyone of this group are in good health. I am working on uncertainty quantification of NSE 2nd degree. I am using umfpack direct solver to solve my system matrix.
In the experiment, if I run 2D mesh with 100k degrees of freedom, I can get my result within 12-13 days of running code. But if I run 3D mesh it takes much longer time than expected.
Any suggestion how I can run 3D mesh in less time period?

Regards,
Md Mahmudul Islam 

Yann Jobic

unread,
Apr 15, 2025, 3:30:52 AM4/15/25
to dea...@googlegroups.com
Hello Md Mahmudul,

I'm running also NSE, and with higher degree of freedom. It's not taking
that much time. In order to understand what is going on, i've got some
questions :
1) are you in debug compilation mode ?
2) it seems that you are solving a linear system with a direct method.
Is there behind an iterative solver somewhere ?
3) you may want to use the TimerOutput class to understand where you are
spending all the time. My guess is that it's not in the direct solver
(umfpack)
Best,
Yann
> --
> The deal.II project is located at http://www.dealii.org/ <http://
> www.dealii.org/>
> For mailing list/forum options, see https://groups.google.com/d/forum/
> dealii?hl=en <https://groups.google.com/d/forum/dealii?hl=en>
> ---
> You received this message because you are subscribed to the Google
> Groups "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to dealii+un...@googlegroups.com
> <mailto:dealii+un...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/dealii/
> CAFAJfA1cdbEZGTtw-zzApjV_hyWWX0nAks1izJYqV6YGSb6dCQ%40mail.gmail.com
> <https://groups.google.com/d/msgid/dealii/CAFAJfA1cdbEZGTtw-
> zzApjV_hyWWX0nAks1izJYqV6YGSb6dCQ%40mail.gmail.com?
> utm_medium=email&utm_source=footer>.

Md Mahmudul Islam

unread,
Apr 15, 2025, 3:35:44 AM4/15/25
to dea...@googlegroups.com

I hope the code is good to go. Sorry I haven't mentioned in my code I have 20 realization numbers and I think it’s taking time here. Because if I reduce the number of realization number its fast.
In that case, should I use an iterative solver?

Md Mahmudul Islam


For mailing list/forum options, see https://groups.google.com/d/forum/dealii?hl=en

---
You received this message because you are subscribed to the Google Groups "deal.II User Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dealii+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/dealii/b91709b4-6060-43f0-84df-58fb88d4c825%40gmail.com.

Yann Jobic

unread,
Apr 15, 2025, 3:47:10 AM4/15/25
to dea...@googlegroups.com
No, the direct solver is good if you have enough memory. The solver may
not be the problem here.
If I understand that realization means one NSE with some particular
parameters, that means that 20 realizations are 20 different NS problems
to solve. Timing your code will help, you can look at
https://www.dealii.org/developer/doxygen/deal.II/step_32.html

Yann
> www.dealii.org/> <http://
> > www.dealii.org/ <http://www.dealii.org/>>
> > For mailing list/forum options, see https://groups.google.com/d/
> forum/ <https://groups.google.com/d/forum/>
> > dealii?hl=en <https://groups.google.com/d/forum/dealii?hl=en
> <https://groups.google.com/d/forum/dealii?hl=en>>
> > ---
> > You received this message because you are subscribed to the Google
> > Groups "deal.II User Group" group.
> > To unsubscribe from this group and stop receiving emails from it,
> send
> > an email to dealii+un...@googlegroups.com
> <mailto:dealii%2Bunsu...@googlegroups.com>
> > <mailto:dealii+un...@googlegroups.com
> <mailto:dealii%2Bunsu...@googlegroups.com>>.
> > To view this discussion visit https://groups.google.com/d/msgid/
> dealii/ <https://groups.google.com/d/msgid/dealii/>
> > CAFAJfA1cdbEZGTtw-
> zzApjV_hyWWX0nAks1izJYqV6YGSb6dCQ%40mail.gmail.com
> <http://40mail.gmail.com>
> > <https://groups.google.com/d/msgid/dealii/CAFAJfA1cdbEZGTtw-
> <https://groups.google.com/d/msgid/dealii/CAFAJfA1cdbEZGTtw->
> > zzApjV_hyWWX0nAks1izJYqV6YGSb6dCQ%40mail.gmail.com
> <http://40mail.gmail.com>?
> > utm_medium=email&utm_source=footer>.
>
> --
> The deal.II project is located at http://www.dealii.org/ <http://
> www.dealii.org/>
> For mailing list/forum options, see https://groups.google.com/d/
> forum/dealii?hl=en <https://groups.google.com/d/forum/dealii?hl=en>
> ---
> You received this message because you are subscribed to the Google
> Groups "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it,
> send an email to dealii+un...@googlegroups.com
> <mailto:dealii%2Bunsu...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/
> dealii/b91709b4-6060-43f0-84df-58fb88d4c825%40gmail.com <https://
> groups.google.com/d/msgid/dealii/
> b91709b4-6060-43f0-84df-58fb88d4c825%40gmail.com>.
>
> --
> The deal.II project is located at http://www.dealii.org/ <http://
> www.dealii.org/>
> For mailing list/forum options, see https://groups.google.com/d/forum/
> dealii?hl=en <https://groups.google.com/d/forum/dealii?hl=en>
> ---
> You received this message because you are subscribed to the Google
> Groups "deal.II User Group" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to dealii+un...@googlegroups.com
> <mailto:dealii+un...@googlegroups.com>.
> To view this discussion visit https://groups.google.com/d/msgid/dealii/
> CAFAJfA27nrwghcHoyROx7EdUf2T%3DRnp86LkzRQ1q3qwgKYcvOg%40mail.gmail.com
> <https://groups.google.com/d/msgid/dealii/
> CAFAJfA27nrwghcHoyROx7EdUf2T%3DRnp86LkzRQ1q3qwgKYcvOg%40mail.gmail.com?
> utm_medium=email&utm_source=footer>.

Reply all
Reply to author
Forward
0 new messages