Fluent 14 processor and memory setting

mosyarus · October 29, 2013, 23:16

HI, I use AMD FX-8120 8 processors and 8gb RAM. When I start Fluent 14 I use parallel processing with 8 processors. When I run Fluent I notice that all my 8 processors running on 100% load and my memory only 20+% used. Is it normal or I can change another setting for better performance.

cdgatti · October 31, 2013, 10:33

Hello,

I have the same computer. The RAM memory usage depends on the complexity of the model that are running, the more complex, more calculations must be done and more RAM to use.

mosyarus · October 31, 2013, 22:11

tq Claudio for the reply,can i have other setting to reduce the calculation time with the same model?

cdgatti · November 1, 2013, 06:50

Making an initial initialization potential flow by FMG, you greatly reduce convergence time (if you're not doing). Another way is to improve the quality of the mesh (skew <0.9). Otherwise reducing relaxation coefficients, and the number Courant.

I hope to be of your help

mosyarus · November 5, 2013, 23:36

Thank you for the suggestions. I hope your suggestions can help me for my simulations. Thanks again

October 29, 2013, 23:16	Fluent 14 processor and memory setting	#1
mosyarus New Member mosyarus Join Date: Oct 2013 Posts: 4 Rep Power: 13	HI, I use AMD FX-8120 8 processors and 8gb RAM. When I start Fluent 14 I use parallel processing with 8 processors. When I run Fluent I notice that all my 8 processors running on 100% load and my memory only 20+% used. Is it normal or I can change another setting for better performance.

October 31, 2013, 10:33		#2
cdgatti New Member Claudio Join Date: Jun 2013 Posts: 8 Rep Power: 13	Hello, I have the same computer. The RAM memory usage depends on the complexity of the model that are running, the more complex, more calculations must be done and more RAM to use.

October 31, 2013, 22:11		#3
mosyarus New Member mosyarus Join Date: Oct 2013 Posts: 4 Rep Power: 13	tq Claudio for the reply,can i have other setting to reduce the calculation time with the same model?

November 1, 2013, 06:50		#4
cdgatti New Member Claudio Join Date: Jun 2013 Posts: 8 Rep Power: 13	Making an initial initialization potential flow by FMG, you greatly reduce convergence time (if you're not doing). Another way is to improve the quality of the mesh (skew <0.9). Otherwise reducing relaxation coefficients, and the number Courant. I hope to be of your help

November 5, 2013, 23:36		#5
mosyarus New Member mosyarus Join Date: Oct 2013 Posts: 4 Rep Power: 13	Thank you for the suggestions. I hope your suggestions can help me for my simulations. Thanks again