# Uneven CPU thermals!

**URL:** <https://community.frame.work/t/uneven-cpu-thermals/55614>\
**Category:** Framework Laptop 16\
**Tags:** framework-laptop-16-amd-7040\
**Created:** [August 3, 2024, 6:55pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614 "2024-08-03T18:55:00Z")\
**Posts on this page:** 20\
**Page:** 73

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 5, 2026, 4:33pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1455 "2026-02-05T16:33:34Z")

</div>

Yeah, there are other tools for windows like MSI Afterburner, but I’ve only ever used those with nvidia cards and idk if they’d work. AMD’s own Adrenaline might let you do it too.

---

<div class="post-metadata">

**Author:** ![PSierra117](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/psierra117/32/25670_2.png) [@PSierra117](https://community.frame.work/u/PSierra117)\
**Post date:** [February 5, 2026, 4:40pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1456 "2026-02-05T16:40:14Z")

</div>

Nope Adrenaline and Afterburner aren’t able to alter the TDP Barrier on the 7700S or i think Mobilechips in general.

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 5, 2026, 4:43pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1457 "2026-02-05T16:43:03Z")

</div>

Well that’s a bummer. At least it will boost to 120W on it’s own by default when it feels like it (I think).

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 5, 2026, 4:44pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1458 "2026-02-05T16:44:43Z")

</div>

Thanks!

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 5, 2026, 4:47pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1459 "2026-02-05T16:47:10Z")

</div>

It can jump to higher power briefly before locking at 100w. I’ve seen the GPU use a maximum of 189w for a brief period.

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 6, 2026, 2:34am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1460 "2026-02-06T02:34:07Z")

</div>

Well, unfortunately for me repasting the dGPU did not resolve my overheating issues ☹  
hotspot temp did drop 10c though, so it at least helped. Just not enough.

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 6, 2026, 2:46am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1461 "2026-02-06T02:46:26Z")

</div>

What kind of fps are you getting? Is it like unstable or just really low?

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 6, 2026, 3:08am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1462 "2026-02-06T03:08:48Z")

</div>

I get a steady ~200 in BF6 for about 25 minutes, and then I get an inconsistent 80 once the power target starts bouncing all over.

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 6, 2026, 3:18am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1463 "2026-02-06T03:18:35Z")

</div>

That’s really weird. That doesn’t seem like a thermal issue. I’m not sure what it could be earlier today I was having a strange stutter in a game and was only able to fix it by removing the battery and letting all the power drain. Not sure if it’s related but sometimes when I play Minecraft with shaders the VRAM fills up and I need to restart the game. I guess see if the VRAM fills up at all but I really am not sure what it could be.

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 6, 2026, 4:34am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1464 "2026-02-06T04:34:40Z")

</div>

Yeah, me neither tbh. I’m running out of components to target chill in the laptop, but I can the problem go away if I blast it with a llano cooling pad underneath.

Update while writing reply: I just reviewed some fresh logs and I think I finally found what’s causing it to throttle though! For some reason, my 780M iGPU started thermal throttling (at 73c) and that same moment is when I ran into my performance issues! Figuring out how to get around that though without blasting it with a big cooler underneath though is going to be a little difficult, considering I’ve already done the PTM sandwhich mod.

Also, idk why it would be getting hot to begin with, other than by being adjacent to other hot things.

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 6, 2026, 4:45am UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1465 "2026-02-06T04:45:18Z")

</div>

If you’re using HWiNFO you can see what the temps are and why something might be throttling. See if it says if it’s from temperature or something else. It also tells you the thermal limit percentage.

---

<div class="post-metadata">

**Author:** ![jared\_kidd](https://avatars.discourse-cdn.com/v4/letter/j/91b2a8/32.png) [@jared\_kidd](https://community.frame.work/u/jared_kidd)\
**Post date:** [February 6, 2026, 2:25pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1466 "2026-02-06T14:25:06Z")

</div>

> [@TechPriestNhyk](#):
>
> For some reason, my 780M iGPU started thermal throttling (at 73c) and that same moment is when I ran into my performance issues!

I doubt it is made to thermal throttle when it hits 73c. What led you to believe that? There are many reasons it may throttle besides heat.

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 6, 2026, 2:56pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1467 "2026-02-06T14:56:10Z")

</div>

I put more information here: [[WIP] An Adventure In Mitigating Heat Soak - #17 by TechPriestNhyk](https://community.frame.work/t/wip-an-adventure-in-mitigating-heat-soak/80255/17)

---

<div class="post-metadata">

**Author:** ![jared\_kidd](https://avatars.discourse-cdn.com/v4/letter/j/91b2a8/32.png) [@jared\_kidd](https://community.frame.work/u/jared_kidd)\
**Post date:** [February 6, 2026, 3:28pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1468 "2026-02-06T15:28:58Z")

</div>

I see lots of graphs, but nothing explaining your method of detecting thermal throttling, or even which throttling flag you are seeing.

You may not be aware, but framework uses the PROCHOT flag as a catch all for any throttling. In fact, my 16 has had that flag triggered since purchase. I can only clear it by sleeping the laptop. But as soon as I disconnect or connect a charger, it’s triggered again. But even with the flag triggered, it’s not thermal throttling and my CPU still hits max package power (if my dGPU is awake, but that’s another topic).

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 6, 2026, 3:45pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1469 "2026-02-06T15:45:40Z")

</div>

I run BF6 until it throttles, this takes almost exactly 25 minutes every time. It’s quite obvious when it thermal throttles, because the framerate drops to 40% of what it was (200fps+ → 80fps). In HWInfo64, the specific field I’m referencing that says the iGPU is thermal throttling is called `Throttle Reason - Thermal (Yes/No)`, this is separate from the prochot flags. In my most recent post on that thread, I show the STAPM limit hitting 100% and power target’s plummeting (when the throttling occurs). I also show plots of the iGPU throwing a thermal throttle flag. I don’t mention PROCHOT flags in my post because during this test the laptop did not throw any PROCHOT flags.

My next steps are to understand more about how _exactly_ STAPM works, and how I can use that knowledge to resolve my performance issues.

---

<div class="post-metadata">

**Author:** ![theNobleOpti](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/thenobleopti/32/46158_2.png) [@theNobleOpti](https://community.frame.work/u/theNobleOpti)\
**Post date:** [February 20, 2026, 11:01pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1470 "2026-02-20T23:01:03Z")

</div>

This is probably going to sound really stupid, but when I replace the liquid metal with PTM7950…

is the grey square with the little dots on it supposed to stay on the CPU cooler when installing? Like, is that part of the cooler, or is it the liquid metal?

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 20, 2026, 11:03pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1471 "2026-02-20T23:03:36Z")

</div>

Part of the cooler, that’s the aluminum shim. There should be solid bits of the “liquid” metal on it though that scrape off. You’ll want to clean that as much as possible.

---

<div class="post-metadata">

**Author:** ![theNobleOpti](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/thenobleopti/32/46158_2.png) [@theNobleOpti](https://community.frame.work/u/theNobleOpti)\
**Post date:** [February 20, 2026, 11:09pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1472 "2026-02-20T23:09:45Z")

</div>

so i uhh, bought a new cooler to fix the thermal issues, and I took the whole aluminum shim off… because I assumed that whole shim was the liquid metal pre-applied to the cooler.

MAN I feel stupid now. That would also explain why it was such a pain to remove, because apparently I wasn’t supposed to.

---

<div class="post-metadata">

**Author:** ![TechPriestNhyk](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/techpriestnhyk/32/27853_2.png) [@TechPriestNhyk](https://community.frame.work/u/TechPriestNhyk)\
**Post date:** [February 20, 2026, 11:14pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1473 "2026-02-20T23:14:56Z")

</div>

Well, I got good news for you then! You’re just a couple steps away from completing the “copper shim” mod and improving the thermal cooling performance 🙂

Essentially, all you need to do for that is to get a PCM pad from LTT (although I imagine thermal paste or other thermal pads would also work), use that thermal pad to attach a 0.8mm copper shim to the heatsink, and then hookup the heatsink as normal 🙂

A proper description of the process is relatively well documented in the thread above, that’s just a poorly explained cliff notes.

TLDR, you’re not screwed 🙂

---

<div class="post-metadata">

**Author:** ![zachary\_stout](https://sea1.discourse-cdn.com/flex001/user_avatar/community.frame.work/zachary_stout/32/31358_2.png) [@zachary\_stout](https://community.frame.work/u/zachary_stout)\
**Post date:** [February 20, 2026, 11:15pm UTC](https://community.frame.work/t/uneven-cpu-thermals/55614/1474 "2026-02-20T23:15:05Z")

</div>

You might as well do the copper shim mod then

[Previous page](https://community.frame.work/t/uneven-cpu-thermals/55614.md?page=72)

[Next page](https://community.frame.work/t/uneven-cpu-thermals/55614.md?page=74)
