track time/memory use statistics reported in RT logs

[DeniseWorthen]

Description

PR #2061 added simplified reporting of compile and run times and memory usage. We don't make use of this valuable information because there is no way to track it.

Solution

There should be a process by which pertinent information can be scraped from the Regression test logs, collated and visually presented so that performance (run time, memory use) for each commit can be tracked.

Alternatives

Related to

[jkbk2004]

We can extend https://github.com/NOAA-EPIC/ufs-dev_logs/tree/main/images to upload benchmark status to EPIC dashboard, https://noaa-epic-dashboard.s3.amazonaws.com/index.html. The dashboard is publicly associable. @FernandoAndrade-NOAA @gspetro-NOAA we can add this item to PI-15 planning.

[DeniseWorthen]

What is more useful is to be able to visualize the change over time, not just the snapshot of the current.

[jkbk2004]

Sure! Those are example figures. No problem to track in time series with differentiated features.