[gmx-users] About getting running time from a Gromacs Run

[Mark Abraham]

On 10/08/2012 3:06 AM, jesmin jahan wrote:
> Dear Gromacs Users,
>
> I am a new user of Gromacs  and I am using gromacs to calculate
> GB-polarization energy .
>
> I have a couple of questions about it. It would be great if someone
> can answer them. Thanks in advance.
>
> Questions:
>
> 1: I got an output like this in log file from a protein simulation:
>
>   R E A L   C Y C L E   A N D   T I M E   A C C O U N T I N G
>
>   Computing:         Nodes     Number     G-Cycles    Seconds     %
> -----------------------------------------------------------------------
>   Comm. coord.          24          1        0.235        0.1     1.0
>   Force                 24          1       17.369        5.2    71.9
>   Wait + Comm. F        24          1        0.258        0.1     1.1
>   Write traj.           24          1        0.296        0.1     1.2
>   Update                24          1        0.005        0.0     0.0
>   Comm. energies        24          2        7.252        2.2    30.0
> -----------------------------------------------------------------------
>   Total                 24                  24.157        7.3   100.0
> -----------------------------------------------------------------------
>
> NOTE: 30 % of the run time was spent communicating energies,
>        you might want to use the -gcom option of mdrun
>
>
> Parallel run - timing based on wallclock.
>
>                 NODE (s)   Real (s)      (%)
>         Time:      0.303      0.303    100.0
>                 (Mnbf/s)   (GFlops)   (ns/day)  (hour/ns)
> Performance:      0.000     54.029      0.285     84.257
>
> Now my question is: What is the time to calculate the GB-polarization
> Energy that is shown in log? Or How to get the time?

It's a side-effect of the "Force" component of the timing breakdown.

> 2. I have a protein with 1.5 million atoms. I tried to run it with
> gromacs on 16 machines each with 12 cores. It generated a segmentation
> fault. Is there any size limit to run in gromacs?

No, but if you run yourself out of memory results are unpredictable. 
Consult your .log and stdout files for clues what is going on.

> 3. Can I only calculate the GB-polarization energy without calculating
> any Forces? How?

http://www.gromacs.org/Documentation/How-tos/Single-Point_Energy

> 4. What are the good tools to get the GB- energy values along with the
> time to calculate the GB-energy?

As above.

Mark