Powertop reports ridiculous powerdraw by wifi

Framework Laptop 13 Linux
1

Which Linux distro are you using? Gentoo
Which kernel are you using? 6.15.3
Which BIOS version are you using? 3.03
Which Framework Laptop 13 model are you using? AMD Ryzen™ AI 300 Series
Mediatek firmware: 20250613_p20250624

The title says it all. Sample output:

Power est. Usage Events/s Category Description
4.16 W 2.0 pkts/s Device Network interface: wlan0 (mt7925e)
2.58 W 3.1% Device Display backlight
1.86 W 100.0% Device USB device: Wireless_Device (MediaTek Inc.)
319 mW 9.6 ms/s 1088.3 Timer tick_nohz_handler

The 2W USB device remains even after I soft rfkill the device.
The reported power draw by the battery doesn’t seem to change by that much IMO, but the sampling isn’t granular enough for me to make that conclusion.

Anyone else seeing this?

2

I’m on FW16 with Gentoo using kernel 6.16-rc3 with Mediatek 7925 and previously Mediatek 7921 and I’ve been seeing the same since almost day one

I think powertop is not accurate, especially when your AC adapter is plugged in

3

Yeah, same, the numbers it reports are wrong

4

I think powertop is not accurate, especially when your AC adapter is plugged in

I agree, all my comments are based on readings while on battery.

It definitely isn’t accurate, or something is buggy elsewhere in the reporting chain: when disabling the laptop screen (connecting an external screen) it keeps on reporting backlight usage at 3W.

5

Yeah, I’m sure it’s powertop ( my opinion ) that’s inaccurate

I find that using powertop while only on battery, the results are more accurate but even then they’re still bum

6

Apart from battery discharge powertop doesn’t have any other sources of information and relies on educated guesses. With modern power management features these are often extremely wrong.

7

I did the unthinkable: I read the manpage. I’ll be playing around with powertop --calibrate tomorrow…