Somewhat surprisingly (but very welcome): these RAM and CPU upgrades, so far, have not materially increased the power demand of my lab.
I put the new RAM in the VM server yesterday (the 25th) and the upgraded CPU/RAM in the lab bench box a few hours ago.
I'm sure if I were to max them out for hours a day, I'd see a difference. But since my compute demands tend to be very bursty, the average power draw is the bigger concern and that seems to be roughly unchanged.
Today was also a new record for solar production. I was inside at the $dayjob lab and not sure what things were like at home, but guessing the same light diffuse-reflector clouds that seem to give the best results with low winter sun.
Instantaneous peak generation was 6.7 kW with a demand of 4.9 kW, exporting 1.8 kW. But that was probably perfect cloud scatter conditions and didn't last long, I was only a net exporter for about 30 minutes.
The total production of 21.5 kWh is still an all-time high (a record that won't stand long, as sunnier weather and longer days approach), meeting the equivalent of about 3.67 hours of my demand. Not bad for the off season and definitely encouraging as far as how much juice I can expect to be getting in the summer.
Andrew Zonenberg
Gabriel Gonzalez