CPU Throttling on Beelink+Photon

I have the Beelink N5095, running latest stable Photonvision on top of Ubuntu Server 22.04.1. I’m currently experimenting with OV9281 and AR0144 USB cameras.

It runs fine for about 70% of the time, but occasionally, for seemingly no reason, the CPU speed will throttle way down, causing the FPS to drop to 5 or so.

I’ve played around with the usual tools, but the actual CPU speed doesn’t seem to respect the software governor, or any minimums that I set:

team696@beelink-696-1:~$ cpufreq-info
cpufrequtils 008: cpufreq-info (C) Dominik Brodowski 2004-2009
Report errors and bugs to [email protected], please.
analyzing CPU 0:
  driver: intel_pstate
  CPUs which run at the same hardware frequency: 0
  CPUs which need to have their frequency coordinated by software: 0
  maximum transition latency: 4294.55 ms.
  hardware limits: 800 MHz - 2.90 GHz
  available cpufreq governors: performance, powersave
  current policy: frequency should be within 2.00 GHz and 2.90 GHz.
                  The governor "performance" may decide which speed to use
                  within this range.
  current CPU frequency is 400 MHz.

I’ve also played around in the BIOS, up to and including turning SpeedStep off, but without apparent effect.

Any ideas?

Can thermals be an issue?

Could be, but it seems weird that @asid61 et al. didn’t have similar issues, considering it’s the same hardware (unless Beelink made some quiet design changes somewhere, which isn’t impossible).

I would try opening your Beelink and see if the termal solution was applied properly if it ends up being a thermal issue.

1 Like

Hm I haven’t seen this even running it for hours. How long does it take for this to be an issue? Could you limit the max frequency to 2GHz to mitigate the heat? Have you measured CPU temp?

I would definitely look at replacing thermal paste.

Well, I’ll be darned-- I opened it up to find that the fan connector was unplugged. One more thing to double-check when you order from Amazon, I guess.

Oddly, I had checked the thermal data (with lm-sensors) and it looked reasonable-- 85 C or so actual, and a limit of 105 C. Maybe the temp sensors aren’t positioned very close to the actual CPU, or something.


Well that’s a unique failure mode, LOL. Glad it was an easy fix!

1 Like

If you look at my guide there’s a recommendation on robustification, but I don’t know if put “glue the fan connector” on there. Thank you for checking this failure mode!