As you can see below my APU is going crazy and all I’m doing is watching a youtube video and surfing the web. I will let my laptop just sit there and it’s like it’s haunted and doing stuff on it’s own because I can hear the fans jack up and then go back down.
I know the liquid metal might keep it hotter than the alternative, but I’d prefer to figure out and fix what’s CAUSING the thermal up and down before replacing anything.
Any help would be greatly appreciated!!!
If you’re still running on the original batche’s LM application, it’s unfortunately probably run off the contact patch at this point and pooled around the die.
You may want to look into pulling off the heatsink and removing the LM to apply normal thermal paste or PTM, it won’t get any better without this I bet.
Thanks for responding. Just so I know I’m understanding you, you’re saying the cooling system could also be the cause of the frequent thermal spikes up and down?
There was a design flaw in the original batches of motherboards that directly lead to thermal issues. Essentially, the liquid metal they used as a thermal interface would leak away from the heatsink over time and lead to a poor thermal connection. If you’ have a liquid metal unit, I would recommend replacing the LM with the PTM material framework provides (which you can request via support ticket).
Your graph looks similar to those I’ve seen for people with the issue (myself included), except your temperatures are still kind of low. When mine had the issue, it would spike to 100c.
If the thermal interface is making poor contact, there’s not a lot of heat transfer between the chip and the heatsink, so your APU is going to heat up with even small loads. The heat simply never leaves fast enough for the temperature to lower.
One think you can do to verify, is install HWInfo and take a look at the CPU’s power usage. I have my machine sitting under ~6-8W on idle, which is pretty normal. You can use HWInfo’s taskbar displays to show temperature and core clocks in the taskbar area. If you see similar low power draws, than it’s most likely the LM deteriorated. If you are actually seeing large CPU power usage, you should look into background processes/window nonsense first.
Check the power consumption and possibly the CPU usage per program on taskmgr/btop
I know how to check cpu usage using task manager, but how do I check power consumption - is that somewhere on Task Manager as well?
In the “processes” menu you can right click the CPU percentage on top and click the “power usage” and “power usage trend”. It’ll give a rough idea, not a precise watt per program
Ok. I done goofed and did too much, or something. Not sure what, but now when I try to play a game, instead of the apu spiking, it PLUMMETS. Steam overlay shows the cpu as rarely being above 5% give or take a few percentage points. See below image.
NOW what do I do‽‽‽
Which program are you using? Can you bring up the APU power there?
I’m using Framework Control, and I don’t think so. I did install Universal x86 Tuning Utility to try and maximize things, but when I realized my cpu behavior changed I uninstalled it. Reinstalled it to see if I can fix what I did, but I’m not sure if that’s even what did it.
Ok so I Uninstalled Universal x86 Tuner Utility, Uninstalled some other random things, rebooted and shut down a few times, double checked BIOS, and my games are playing well like they used to before I went overboard and seemingly installed and changed too much and got in over my head. So I’m back to square one with the spiky apu temp line. I think I’ll take it easy on more software or more settings changes until I replace the liquid metal.
You are better off without it. It uses winring0 which is a nuclear catastrophic vulnerability at this point. It was always goofy, always crap, when it was on mine as well.
also as everyone else above, run a benchmark. cinemark. if your temps go into orbit while performance falls off a cliff then you might want to pry off the heatsink to clean off the spent, leaking LM before repasting with frameworks replacement if you submit a claim.
Thermalgrizzly Phasesheet works as well because they guarantee it is PTM7950 (is that right?) if you are impatient. I say that because purchasing from Amazon or the like is not a guarantee of the correct thermal material.
So as an update after looking at UXTU again, I realized that I probably fucked up my pc temporarily because with so many settings I had no point of reference as I’m new to overclocking. However, I then came across a piece of software called Saku Overclock. It has many of the same features as UXTU, except there are RECOMMENDATIONS for each setting based on if you want more power saving of performance. It’s a little buggy, but I set up a custom profile and set it up, and it has reduced my fan noise, allowed me to overclock a bit, reduce the APU temp, and monitor resources. It’s a great little app that’s still a work in progress but it worked for my gaming this evening fantastically. I still will switch out the liquid metal at some point, but this helps a lot.
you want to take a step back. you have too many variables in too many applications. who knows what is actually happening, fix your thermal problems first if that is the true cause, then you can play with more settings. with proper backups in place so you can fix fuckups faster
Even if the cooling is ok, what makes the OP’s CPU temperature lower when playing video games?
I don’t know what I did there it was bizarre.
Fair point. The rabbit holes I go down on my days off…
