enjoyable simulation and time analysis
Clicked A Few Times
1:57:56 AM PST - Mon, Mar 4th 2013
Hi everyone,

I have run an QMMM optimization with NWChem? the qmlink region contains 185 atoms and MM region (mm_solute and solvent) contains about 50,000 atoms. It is completed within 30 hours. I am very enjoyable this simulation, more and more like NWChem program, and I wish give my sincere thanks to the developers.

Now I have a little question about time analysis. At the end of main output file, we can see the total time:

Total times  cpu:    99739.1s     wall:    94377.0s

At the end of optimization output of mm region, we can the following line:

                   Number of calls         Average times per call



                                            Wall            CPU



MD step                  254              2131.533       2249.503

I think the total time 99739.1 is ONLY for 1 core. I use 16 cores in this simulation. However, the mm optimization time 2131.55 should for 16 cores. Am I right?

Thanks in advance.

Jingbo

Clicked A Few Times
10:37:55 AM PDT - Wed, Jun 19th 2013
We are glad to hear you are enjoying the code.
Total CPU/wall time should refer to a total execution time of your 16 core run, not just one core. I need to check regarding time statistics for mm optimization, there may be some issues due to the nature of QM/MM run.

Marat

Just Got Here
2:37:32 PM PST - Fri, Dec 27th 2013
share input
Can you please share your input. Just to get a scheme for how to do mine. So far I have had no luck. Please also include your prepare section to see how you created the topology and restart files if you had to.

Thank you
Philipa


Forum >> NWChem's corner >> QMMM