Emergency Request

Ali T · December 12, 2013, 02:52

Dear all friends

How can we run NUMECA fine\turbo in parallel computing? (For CFD and Optimization!)

I have NUMECA fine\turbo 8.9.1.

__________
Best Wishes

colinda1 · January 16, 2014, 10:42

You can easily define a parallel computation through the TaskManager (accessible through the Modules menu of FINE/Turbo). You can define, launch and monitor your parallel computation directly from the TaskManager or alternatively you can use the TaskManager to prepare the files needed to launch in batch.

alederiaz · January 23, 2014, 09:43

Is it possible to create a more balanced partition between all the processors or we can only allocate numbers of blocks to each processor?

Thank you

colinda1 · January 23, 2014, 09:53

Sure, in FINE/Turbo v9 you can use parallel partitioning to have a better balance over all the processors.
In the Task Manager simply select on the Tasks definition page: "Parallel Partitioned Computation"

alederiaz · January 29, 2014, 06:30

Should I select 'parallel computation' after starting the solver?
If I want to do it before I can lot select the computation from the list. If i do it later it do not use all slave processor selected
Probably I missed a step.

Thank you

colinda1 · February 6, 2014, 03:38

In the FINE TaskManager you should first select the *.run file of your computation. Then select "Parallel Computation" and modify the "Parallel Settings" all before starting the solver.

December 12, 2013, 02:52	Emergency Request (Multi prossecor)	#1
Ali T New Member Join Date: Nov 2013 Posts: 2 Rep Power: 0	Dear all friends How can we run NUMECA fine\turbo in parallel computing? (For CFD and Optimization!) I have NUMECA fine\turbo 8.9.1. __________ Best Wishes Last edited by Ali T; December 12, 2013 at 07:17.

January 29, 2014, 06:30		#5
alederiaz New Member Deneriaz Alphonse Join Date: Nov 2013 Posts: 7 Rep Power: 12	Should I select 'parallel computation' after starting the solver? If I want to do it before I can lot select the computation from the list. If i do it later it do not use all slave processor selected Probably I missed a step. Thank you Last edited by alederiaz; January 29, 2014 at 12:35.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Mesquite - Adaptive mesh refinement / coarsening?	philippose	OpenFOAM Running, Solving & CFD	94	January 27, 2016 09:40
[OpenFOAM] Saving ParaFoam views and case	sail	ParaView	9	November 25, 2011 15:46
Problem with paraFoam	ata	OpenFOAM	14	November 30, 2009 04:23
Installation OF1.5-dev	ttdtud	OpenFOAM Installation	46	May 5, 2009 02:32
[OpenFOAM] Xwindows crash with paraview save	srinath	ParaView	1	October 15, 2008 09:37

January 16, 2014, 10:42		#2
colinda1 Senior Member Colinda Join Date: May 2012 Location: Brussels Posts: 151 Rep Power: 13	You can easily define a parallel computation through the TaskManager (accessible through the Modules menu of FINE/Turbo). You can define, launch and monitor your parallel computation directly from the TaskManager or alternatively you can use the TaskManager to prepare the files needed to launch in batch.

January 23, 2014, 09:43		#3
alederiaz New Member Deneriaz Alphonse Join Date: Nov 2013 Posts: 7 Rep Power: 12	Is it possible to create a more balanced partition between all the processors or we can only allocate numbers of blocks to each processor? Thank you

January 23, 2014, 09:53		#4
colinda1 Senior Member Colinda Join Date: May 2012 Location: Brussels Posts: 151 Rep Power: 13	Sure, in FINE/Turbo v9 you can use parallel partitioning to have a better balance over all the processors. In the Task Manager simply select on the Tasks definition page: "Parallel Partitioned Computation"

February 6, 2014, 03:38		#6
colinda1 Senior Member Colinda Join Date: May 2012 Location: Brussels Posts: 151 Rep Power: 13	In the FINE TaskManager you should first select the *.run file of your computation. Then select "Parallel Computation" and modify the "Parallel Settings" all before starting the solver.