How to evaluate the calculation time of CFX case
Dear all,
More haste ,less speed. If I can get the calculation time about CFX cases, I will work more efficient. Quote:
|
I do not understand your question? What are you trying to achieve? Have you looked up definitions of simulation time, CPU time, etc?
|
Take an example ,I defined a case(def) and than how long is it to get its results(res)?Thank you for your kindness
|
In the output file there are two groups of timing statistics reported.
At the very end of the output file is the total time. This is the time taken for setting up, doing the simulation and packing up at the end. Wall clock and CPU time are reported. A second time is reported after the iterations but before the final balances and residuals is written. This is the wall clock and CPU time taken for just the simulation iterations. When you do benchmarking you will find the setting up and packing up are often very hard drive intensive and do not scale in parallel simulations, as most of the work is done on the master node. But the simulation time should be only CPU time and should scale with the number of parallel nodes. So most multi-processor benchmarking is done using the simulation time, not the total time. |
All times are GMT -4. The time now is 06:03. |