4 ms·
Yeah performance is better in most benchmarks but the power consumption is always double, and given we are in a home server context it's a pretty important chec
by PedroBatista 3y ago
Yeah performance is better in most benchmarks but the power consumption is always double, and given we are in a home server context it's a pretty important checkbox ( even if at this level it could be more emotional than anything else )
In the last few years more is demanded from these ARM boards, so is the price demanded from who makes them. Some x86 boards are right there in some of that territory..
- ekianjo 3y agoARM boards consume more and more power generation after generation so its probably be going parity power performance profile in the end. Also the key problem for ARM boards is piss poor Linux support beyond release so x86 always wins there
- ThrowawayR2 3y agoDoubling an idle power of 5 watts is 10 watts. Even at $0.25 kWh, that's the difference between ~$11 and ~$22 per year. That's not nothing but it's not really a big deal either if the advantages of being x86 apply to your situation.
- mosselman 3y agoSince the question of which hardware is better, usually same performance at double the watts makes a cpu worse. Also you are assuming infinite power availability. If you are off the grid and have power from solar or wind, then lower wattage might be preferable.
- justsomehnguy 3y agoI did not go to bar yesterday. I saved way more than $11. If this is not something which would run on a solar panel back in the woods then additional +5W is nothing in the common household.
- theodric 3y agoI'm saving infinite money by never going to the bar at all! I'm wealthy beyond dreams! In reality, one 5W bump isn't going to move the needle, but you only have to make that concession a few times before you end up adding an appreciable amount to your power bill. I'm paying €0.39/kWh for power. When I added Starlink (57-75W constant draw, equating to at least €12/mo on the bill) I went looking for things to unplug to wash that increase out. Turns out my power consumption monitoring plugs were using about 15W on their own, and having too many wireless access points was adding another 45W. That's all been sorted out, and I'm more cautious of suffering death by a thousand cuts now.