Too big units, comparing to just finished ones

[Lazvon]

Well, my solar electrons shove all the other electrons along the path forward… but if I didn’t pay they wouldn’t even be allowed to shove… or something like that.

[appepi]

...

The newest LAR Systems dark mode Chrome extension for Folding@Home has a rudimentary time-of-day setting which you can use from the browser, instead of setting up scripts in the OS. This time-of-day setting issues a "start" and "finish" command at the specified time, but if the WU is long-running, then it might never actually pause; it's more tuned to finish shorter-running WUs.

Thanks gunnare, I have finally gotten around to trying the LAR systems dark mode extension, with start/finish times set to align with "off-peak" electricity rates (10 pm - 7am). I haven't worked out what my average job-time is, but it is rarely much more than 4 hours, and on average the last job starting before 7 am would be half-way through. Thus if the command "finish up and stop" is sent at 7am my average "worst-case" scenario is that the job continues to 9am and I have used 2 extra hours at "shoulder" rates (equivalent of about 3 hours at off peak rates), and the actual worst case would be 4 extra hours (equivalent of 6 in costs) unless I intervene manually. Meanwhile other GPUs will have finished earlier, but the Z600 or Z800 or Z440 boxes will still be consuming power on idle and producing zip until I wake up and manually switch them off. This seems to work reasonably well, (as long as I don't sleep in too late).

[appepi]

Team appepi 3rd Annual Report

Somewhere in the Forum I already published a 3-year comparison of my average costs per MPoint since I started trying to contain Folding costs in a systematic way, but at that stage the third 52-week period wasn't complete, and anyway I can't find it again. So I will add it here, where the thread includes a number of the helpful "power management" contributions from others that I have implemented.

I note that the "Delta$" costs I refer to here are from a real-world experiment counting ADDITIONAL costs relative to unmanaged non-FaH usage over a 2 week period in May 2022. Thus it captures the overall effect of FaH and other things, and particularly the benefits of paying attention to electricity use. "Management" thus generates free electricity for folding when I switch off unneeded lights, or avoid using power during peak cost times on my Smartmeter (2pm - 8pm) and so on. It has also had the benefit of seeking out and switching to a more suitable plan from my supplier, changes to GPU-only folding, etc. Contrariwise, it suffers from renovation demands that didn't happen in the reference period- eg paint stripping, rendering [millenials please refer to https://en.wikipedia.org/wiki/Cement_render]. All costs are in AU$ of the time, so it also suffers from inflation.

Year 1: (Weeks 001-052) 19 Jun 2022 - 17 Jun 2023. 1,192 MPoints with Delta$ = $405.52 = $0.34 per MP (16.8% of total electricity costs of $46/week)
Year 2: (Weeks 053-104) 18 Jun 2023 - 15 Jun 2024. 0,706 Mpoints with Delta$ = $204.96 = $0.29 per MP (9.4% of total electricity costs of $42/week)
Year 3: (Weeks 105-156) 16 Jun 2024 - 14 Jun 2025. 1,271 Mpoints with Delta$ = $124.06 = $0.10 per MP (6.0% of total electricity costs of of $40/week)

I assume that FaHfolk will understand that keeping das blinkenlights shining is its own reward, so I don't include capital costs.