CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > General Forums > Hardware

Workstation specs for unsteady aerodynamics

Register Blogs Community New Posts Updated Threads Search

Like Tree1Likes
  • 1 Post By flotus1

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   September 8, 2022, 03:28
Default Workstation specs for unsteady aerodynamics
  #1
New Member
 
Daniel
Join Date: Jun 2020
Posts: 3
Rep Power: 5
Mahd is on a distinguished road
What would be an appropriate workstation to simulate 3D unsteady dynamic mesh computational aerodynamics (compressible/incompressible)
problems with about 80 million cell domain using Ansys Fluent? Also it would be better if the same workstation has capabilities to simulate fluid structure interactions.

Actually I need to submit some proposals (budget vs solving time). Hence, the budget has not been determined yet, but the simulation has to be done in an acceptable time (within few days to a week maximum).

If anyone has an idea about the core*hours (@3GHz) for such simulations that will be so helpful.

Any help will be appreciated
Mahd is offline   Reply With Quote

Old   September 8, 2022, 17:07
Default
  #2
Super Moderator
 
flotus1's Avatar
 
Alex
Join Date: Jun 2012
Location: Germany
Posts: 3,399
Rep Power: 46
flotus1 has a spectacular aura aboutflotus1 has a spectacular aura about
It will be nearly impossible to estimate how long your simulations will take on any given hardware. There are just too many variables involved.

From your requirements thouch I would say you should be shopping in the 20k-25k€ price range. Transient, dynamic mesh and 80 million cells are fairly challenging. I would not go below two Milan-X CPUs and probably even 512GB of RAM. 256GB might be enough, but it could be pretty close. I'd rather be conservative here, doubling memory won't add too much to the total cost.
And please check how many threads your Fluent licenses are good for.

The other option is always going to a cluster. You can get even faster solver times than on a single workstation, and your upfront costs are much lower. Again, licensing is a factor here. There is no use for thousands of cores if your licenses only allow e.g. 128 threads.
Mahd likes this.
flotus1 is offline   Reply With Quote

Old   September 9, 2022, 01:17
Default
  #3
Senior Member
 
Will Kernkamp
Join Date: Jun 2014
Posts: 308
Rep Power: 12
wkernkamp is on a distinguished road
Mahd,


Perhaps your engineers have a sample problem that they could run on an AWS server. That should give some insight into convergence and run time.


The run time can be adjusted for a difference in processors, cores, cache, and memory for your system configuration.
wkernkamp is offline   Reply With Quote

Old   September 9, 2022, 01:56
Default
  #4
New Member
 
Daniel
Join Date: Jun 2020
Posts: 3
Rep Power: 5
Mahd is on a distinguished road
@Alex
Thank you for your reply, that was helpful.
The number of threads for the Fluent licenses will be studied based on our needs.
So do you have suggestions for the cluster specs?


@wkernkamp
Thanks. The current resources are quite poor, extrapolating from small domain to a large one might work for determining some of the configurations like RAM, but for the solving time, I think it's difficult to be estimated by linear extrapolations.
Mahd is offline   Reply With Quote

Reply


Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On


Similar Threads
Thread Thread Starter Forum Replies Last Post
Specs for a good workstation gcoopermax Hardware 5 April 28, 2021 10:36
Suggestions for StarCCM+ Workstation configuration ifil Hardware 15 October 30, 2018 05:09
Another Workstation spec questions.... jeanhyuk Hardware 4 July 10, 2018 05:37
Abysmal performance of 64 cores opteron based workstation for CFD Fauster Hardware 8 June 4, 2018 10:51
PC vs. Workstation Tim Franke Main CFD Forum 5 September 29, 1999 15:01


All times are GMT -4. The time now is 23:17.