11th gen overheating when plugged in

I have an early batch of the 11th gen models. i have i7-1185G7 motherboard. i used to have battery drain issues. i recently updated the bios and that seemed to fix it. over the last few days a new issue has come up. when plugged in and near full charge the frequency will spike on all cores (i even saw 5.5ghz on one core in hwmonitor) and the temperature will go to high 90s and the fan will turn on. this isn’t a response to a load i a putting on the machine. i only have chrome and discord open. i only have windows built in anti-virus so its not something like that running in the background. I have my battery charge limit set to 90 in the bios to extend my batteries life. When it first happened i had the bios set to non-turbo performance mode. it got so hot that it shut down the computer. today i changed it to max battery for the performance mode in bios and it still happened. idk what could be causing this. i haven’t dropped it so i don’t think its something like the cooler not making good contact.

any suggestions would be helpful.

What made you think it was overheating?

well when it shut off unexpectedly while looking at a pdf. it was very hot to the touch. then when i rebooted it was at 99C. i didn’t put a new load on the system that would cause that. it just suddenly spiked in temp then shut off. today it started to happen then i unplugged the charger and it stopped boosting the cpu unnecessarily. it doesn;t happen all the time just randomly when plugged into power it decides i want to run at 5Ghz all core now and push 38W through the cpu. idk why it is doing this.

do you hear the fan turning on?

This is a software-driven matter.

Windows update or something.

This is likely a hardware issue.

My laptop (Windows 11) used to continuously spin up its fans and get very hot. It’s never died due to heat yet. I tracked it down to some Windows Services which were essentially just sending telemetry information to Microsoft. Did a load of Regedits to kills them off and now the laptop runs nice a quiet (unless ofcourse I am running a few “Electron” apps).

@SuperCookie this reminds me of my older NUC, which after a year and half of working like a champ started to die (shutdown) of heat last summer. Sounds like your assessment is very close. Thermal paste might be losing it’s efficacy in transferring heat and bad Windows Telemetry code is spooking the CPU.
I am assuming you can hear the fan spin up nice and loud when it starts to get hot.

Switch off the CPU boost in bios and see how it goes then.

yes the fan came on to try an cool it. for some reason even though it was at tjmax something was telling the cpu to keep boosting.

i don’t think its hardware because temps are normal 99% of the time. i tried doing R23 multicore on battery it set the temp limit to ~80C and it ran fine. when i plugged it in to do the same test it when up to 100C and it would thermal throttle, drop the frequency on that core to like 400Mhz, then because it was at 97C and not 100C it would boost it again and the cycle repeats.

i can try a windows update later. i can do more testing(software and hardware) next week after my lab report and mid term are over.

Software shouldn’t cause the laptop to “shut off unexpectedly”.

Thermal throttling is expected with the Framework laptop…but…

…but shouldn’t be clocked down to 400MHz.

the software isn’t shutting it off unexpectedly, rather it was the over heating protection that did. from running other benchmarks there does not seem to be a problem with the cooler. the temps are normal under normal loads on and off battery.

clocking down to 400Mhz is normal behavior for intel cpus with c-step and such enabled. when nothing is happening the frequency is dropped to saved power.

the drop i was talking about was the system throttling itself because of temperature when plugged in and running a benchmark. when the benchmark starts it boosts the frequency. on my cpu its between 4.5ghz and 4.8ghz. when it keeps running like that it will hit tjmax. that is expected for this form factor. when it thermal throttles it drops the frequency. also normal.

the problem i had is how the system acted when under a low load. the first time it happened i had my laptop plugged in and it was close to my battery charge limit of 90%. i was just doing hw with a pdf open. while i was doing this it decided to start boosting as if someone was running prime95. however there was no load. decpite that it kept boosting the cpu even after it hit tjmax. it didn’t thermal throttle like it should have. so the temps went up and the protection circuitry kicked in and shut down the laptop.

i didn’t have hwmonitor on when this happened so that is my best guess to what happened. all i know is that i put a very small load on my laptop (viewing a pdf) and it overheated. when i turned it back on and opened hwmonitor the cpu was at 99C. but after that temps returned to normal. the cooler was able to do what it was suppose to do.

the second time it happened i had hwmonitor open and was able to see it boosting unnecessarily. so i unplugged the charger and it stopped.

there is something wrong with the boosting logic when the laptop is plugged and near a full charge. i have not been able to test anything yet. i will as soon as a i can.

hopefully i explained what happened better.

Have you tried the behaviour with boosting switched off?

i haven’t done any testing yet. the problem cropped up randomly so it might have been a 1 time thing that can’t be reproduced

Nope. Not for 11th gen.

What’s the “something wrong” that you’ve observed? …as opposed to what?

It is expected to boost to 100c-ish.

You can refer to some of the charts and graphs from this thread:

from intel’s 11th gen data sheet. Low Frequency Mode (LFM) for UP3 class chips (which the 1185G7 is a part of) is 400Mhz. it is normal.

The “something wrong” is the processor boosting while it should have been at basically idle and causing the over heating protection to trip and shutting down the laptop. it has only happened twice, with me stopping it the second time by unplugging the charger.

I’m saying, the 0.4GHz LFM is not entered just because “nothing is happening”.

i.e. an Idle UP3 doesn’t drop to 0.4GHz (something else is the trigger here)…at least not when implemented on the Framework Laptop.