analyze timings

[eclare108213]

Update the timings plot for v6.

https://github.com/CICE-Consortium/CICE/blob/2f79cb314051ece262b7497dd197831559d66ee7/doc/source/user_guide/ug_implementation.rst

Figure :ref:fig-timings

[eclare108213]

Do any of the modeling centers running CICE output timings in your regular runs? If so, please paste them here.

[DeniseWorthen]

@eclare108213 Are you referring to the information in the ice_diag.d under "Timing information"? We do have that information for all our regular tests (1deg, 1/2deg and 1/4 deg tripole) but we keep no statistics, so the numbers would be a single case.

[eclare108213]

Yes, that's what I'm after. If you have a bunch of them and they are easy to collate, could you put them in a document or spreadsheet, and post it here? Otherwise one example per grid would be a great place to start. Thanks!

[DeniseWorthen]

Here's a set from a recent run on platform Hera. They're for different run lengths. The 1deg is 24 hrs, the 1/2deg is 30hrs and the 1/4deg is only 6 hours.

It isn't easy to put them into a spreadsheet because of they way they're written. Having the min,mean,max on one line would be a lot easier.

hera.timings.1deg.txt
hera.timings.p5deg.txt
hera.timings.p25deg.txt

[eclare108213]

Thanks @DeniseWorthen, this is very helpful. Please, would you post one of the full ice_diag.d files? I'll grab some configuration specifics (like ncat, nilyr, etc) from it.

[DeniseWorthen]

Here is the 1/4deg ice_diag file. All the configurations use the same basic configuration.

ice_diag.d.p25deg.txt