[AshamedCaptain]

> * A Pi will sit much lower in total power consumption than almost any used PCs if both are doing effectively nothing (ie - simple, spiky tasks like filtering DNS, serving static content from RAM, etc.). You need to be doing something with the system before a PC server comes out ahead, and most people using a Pi as a home server... aren't.

That's not necessarily true since the Pis are particularly terrible at idle power consumption. E.g. the "power off" state consumption shown in the article is actually higher than the idle consumption of some low-power Atom/Celeron x86 chips. The Pi is just terrible at power management.

[suprjami]

Citation needed?

Pi 2 and 3 typically sit at 200 mAh and 230 mAh, Pi 4 is not far away. Zero 2W can go down to 96 mAh.

https://www.jeffgeerling.com/blogs/jeff-geerling/raspberry-p...

https://www.jeffgeerling.com/blog/2021/disabling-cores-reduc...

I don't see any x86 system approaching those numbers.

[ssl-3]

FYI: mA, not mAh.

milliAmps (mA): This is a measure of current flow. Think of it like the flow rate of water through a hose. It's the same kind of unit as something like liters-per-minute is: Where mA is a measure of electrical current flow through a wire (or a device or whatever), liters-per-minute is similarly a measure of the flow of water through a pipe (or consumed, or whatever).

milliAmp-hours (mAh): This measures how much current a something like a battery can supply over time. Imagine it as the total volume of water a hose can deliver if left on for an hour. If a battery is rated at 1000 mAh, it means it can provide a current of 1000 milliamps for one hour, or 500 milliamps for two hours, and so on. To use another water analogy, mAh is like describing the volume of water that is inside of a bucket.

The terms are not interchangeable.

[suprjami]

I understand these things.

If a Raspberry Pi draws 200 mA for one hour, I think it's reasonable to say it has drawn 200 mAh.

[ssl-3]

It is not reasonable, because you did not specify enough information for the reader to draw that conclusion.

> Pi 2 and 3 typically sit at 200 mAh and 230 mAh

200mAh? Over the course of an hour? A day? A fortnight? Just one time, to kickstart the internal perpetual particle accelerator and continue infinitely without additional input? The phraseology used could have specified this information, but it did not do so.

One may wish these units would mean something other than what they do mean, but reality is simply not that way.

We aren't generally free to invent our own scientific nomenclature, or at least we aren't free to do so if effective and meaningful communication is a goal.

[AshamedCaptain]

In TFA they are quoting a whopping 3W on idle for the RPi5 (just search for "idle power"). I have a Celeron system that idles _measured at the wall_ at 1.8W, including 8GB of LPDDR3 and a 4TB SATA SSD. Since I measure at the wall, I'm including transformer losses, which is not usually the case if you just measure current at the USB level as you seem to be doing.

They are also saying that the RPi5 in the default power off state (which is not even a real power off) it stays at 2W.

This matches my own results from the RPi4, where I had difficulty getting it to idle at less than 3W at the wall. While my x86 result is _out of the box_ with an standard openSUSE install.

The (desktop) RPi devices are just TERRIBLE. Cheap, small, have multiple GPIOs, but terrible power-wise. The µc RPis are another story.

[nick238]

What are the keywords to search for to find these Celerons? I assume they're ye'olde business pseudo-light terminals? Laptops are not bad, but desktop machines have easier connectivity.

[AshamedCaptain]

I have a ASUS PN40; on this thread they also mention some N100-based ones which are more powerful. Distrust any result taken with Windows.

[bri3d]

I might just buy one now because I find your wall-power results extremely surprising. My results with basically any mini-PC have been similar between Linux and Windows at minimum idle power; the Windows installs jump up off idle more often but I haven't seen a significant difference in the troughs. The Energy* and other reviews of PN40 indicate ~5W idle, which is more in line with NUCs and everything else I've ever seen from a mini-PC.

[AshamedCaptain]

The difference between Linux and Windows is almost night and day. On my monster workstation, for example, I have managed to idle it down to 16W (or around 24W with a dGPU. it has alder lake, 6 SSDs in softraid and 128GB of RAM). I never managed to do anything less than 40W in Windows.

> 5W which is more in line with NUCs and everything else I've ever seen from a mini-PC.

Even on this very thread you have been quoted lower numbers. Just search around.