Run calculix multithreaded

[flachyjoe]

Hi,
when the env variable OMP_NUM_THREADS is set to the number of CPU, CCX calculation is multithreaded.
BUT it's not the case when launched from FreeCAD. Is python able to launch a multithread process ?

[UR_]

Please set preferences to:

Preferences.png (89.61 KiB) Viewed 4823 times

and you will get:

Resource Monitor.png (27.55 KiB) Viewed 4823 times

@bernd:
# of threads and # of cpu's is mixed up, perhaps should be clarified

edit:
to unleash ccx, you should try pardiso build https://xa.yimg.com/df/CALCULIX/calculi ... e=download
I think you have to be registered for download https://groups.yahoo.com/neo/groups/CALCULIX/info



done with:
OS: Windows 10
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.14426 (Git)
Build type: Release
Branch: master
Hash: 3927a2ed8b009ab4dfdca53fbebf0060193b67b4
Python version: 3.6.6
Qt version: 5.6.2
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: German/Germany (de_DE)

[bernd]

Do you mean the text in prefs should be: „Number of threads to use (Spooles only)” ?

[schupin]

I've got a related question :

On windows, you've to be admin to set the OMP_NUM_THREADS environnement variable.
If you don't change this value but you do it in the FC FEM preferences (as a non-admin), does it work ?

(sorry, I can't test myself, my strange problem isn't solved yet ...)

[UR_]

On windows, you've to be admin to set the OMP_NUM_THREADS environnement variable.

on windows 10 environment variables are divided into two sections:
- system variables (only editable for admins)
- user account variables (editable for non admins too)
So there are no problems setting OMP_NUM_THREADS

[UR_]

Do you mean the text in prefs should be: „Number of threads to use (Spooles only)” ?

to be as precise as could be, what's about:

Set OMP_NUM_THREADS to:

BTW,
- other solvers (e.g. pardiso) are using this env var too (not only spooles)
- there are many additional env vars, which are used by ccx to adjust multi threading on several tasks
OMP_NUM THREADS
CCX_NPROC_STIFFNESS
CCX_NPROC_EQUATION_SOLVER
CCX_NPROC_RESULTS
CCX_NPROC_VIEWFACTOR
CCX_NPROC_CFD
CCX_NPROC_BIOTSAVART

[flachyjoe]

Please set preferences

Thank you !!

[Kailee71]

Hi all,

I'm running the latest Freecad 0.18.4.980 on Windows 10, two Xeon 5670's. I've been trying to get Calculix to run analyses multithreaded but I'm having no luck (yes, I'm using Spooles). I've tried setting the environment variables as well as changing the prefs, but no combination of these allows Calculix to run on more than one core.

Anyone have an idea what I might be overlooking to get all my cores to share the work?


Cheers,

Kailee.

[Kunda1]

Hi all,

I'm running the latest Freecad 0.18.4.980 on Windows 10, two Xeon 5670's.

Sidenote: the FEM subforum sticky note says to run the most up to date version of 0.19_pre
https://github.com/FreeCAD/FreeCAD/rele ... g/0.19_pre

[Kailee71]

Hi all,

I'm running the latest Freecad 0.18.4.980 on Windows 10, two Xeon 5670's.

Sidenote: the FEM subforum sticky note says to run the most up to date version of 0.19_pre
https://github.com/FreeCAD/FreeCAD/rele ... g/0.19_pre

Hi Thanks for your reply - unfortunately this made no difference; any other pointers?

Code: Select all

0.0: Run CalculiX...
0.0: Starting CalculiX...
0.1: CalculiX is running...
70.2: CalculiX stopped.
70.2: ************************************************************ 
CalculiX Version 2.16, Copyright(C) 1998-2019 Guido Dhondt 
CalculiX comes with ABSOLUTELY NO WARRANTY. This is free 
software, and you are welcome to redistribute it under 
certain conditions, see gpl.htm 
************************************************************ 
You are using an executable made on Sun, Jan 19, 2020 10:23:46 PM 
The numbers below are estimated upper bounds 
number of: 
nodes: 39760 
elements: 26891 
one-dimensional elements: 0 
two-dimensional elements: 0 
integration points per element: 4 
degrees of freedom per node: 3 
layers per element: 1 
distributed facial loads: 0 
distributed volumetric loads: 0 
concentrated loads: 5414 
single point constraints: 5046 
multiple point constraints: 1 
terms in all multiple point constraints: 1 
tie constraints: 0 
dependent nodes tied by cyclic constraints: 0 
dependent nodes in pre-tension constraints: 0 
sets: 5 
terms in all sets: 155590 
materials: 1 
constants per material and temperature: 2 
temperature points per material: 1 
plastic data points per material: 0 
orientations: 0 
amplitudes: 2 
data points in all amplitudes: 2 
print requests: 1 
transformations: 0 
property cards: 0 
*INFO reading *STEP: nonlinear geometric 
effects are turned on 
*WARNING reading *STATIC: a nonlinear analysis is requested 
but no time increment nor step is specified 
the defaults (1,1) are used 
STEP 1 
Static analysis was selected 
Newton-Raphson iterative procedure is active 
Nonlinear geometric effects are taken into account 
Decascading the MPC's 
Determining the structure of the matrix: 
number of equations 
114234 
number of nonzero lower triangular matrix elements 
4335234 
increment 1 attempt 1 
increment size= 1.000000e+00 
sum of previous increments=0.000000e+00 
actual step time=1.000000e+00 
actual total time=1.000000e+00 
iteration 1 
Using up to 1 cpu(s) for the stress calculation. 
Using up to 1 cpu(s) for the symmetric stiffness/mass contributions. 
Factoring the system of equations using the symmetric spooles solver 
Using 1 cpu for spooles. 
Using up to 1 cpu(s) for the stress calculation. 
average force= 0.022719 
time avg. forc= 0.022719 
largest residual force= 0.000579 in node 11397 and dof 2 
largest increment of disp= 6.555358e-03 
largest correction to disp= 6.555358e-03 in node 16 and dof 2 
no convergence 
iteration 2 
Using up to 1 cpu(s) for the symmetric stiffness/mass contributions. 
Factoring the system of equations using the symmetric spooles solver 
Using 1 cpu for spooles. 
Using up to 1 cpu(s) for the stress calculation. 
average force= 0.022722 
time avg. forc= 0.022722 
largest residual force= 0.000000 in node 11393 and dof 2 
largest increment of disp= 6.555358e-03 
largest correction to disp= 5.515313e-07 in node 16 and dof 2 
convergence 
the increment size exceeds the remainder of the step and is decreased to 0.000000e+00 
Using up to 1 cpu(s) for the stress calculation. 
Job finished
70.2: CalculiX done without error!
70.2: Loading result sets...