Question GPU laptop overheating right when I start a game

[futureal]

i recently change the thermal paste of my laptop and the GPU temp start to act weirdly. I have a gigabyte g5 (rtx 3060) and now when a start any game the temps go to 86, 87...I reapplied the paste multiple times using a fair amount but still the same behavior.
I also want to mention that the GPU fan is also acting weird, when the temp rises it doesn't increase the rpm quickly. I tested the fans at max speed before starting a game, but the behavior is the same regarding the temps.
I am going crazy it this should I admit defeat and start looking at black friday deals 😂

 

[Aeacus]

i recently change the thermal paste of my laptop

Why?
What was wrong with factory applied thermal compound?

I reapplied the paste multiple times using a fair amount but still the same behavior.

Then change the paste? Since not all thermal pastes are created equal.

~9 years ago, TH made a roundup of thermal pastes, comparing loads of different ones,
link: https://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-19.html

 

[futureal]

Why?
What was wrong with factory applied thermal compound?



Then change the paste? Since not all thermal pastes are created equal.

~9 years ago, TH made a roundup of thermal pastes, comparing loads of different ones,
link: https://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-19.html

I try different ones already (hydronaut, kryonuat). I suspect the problem is somehow connected to the temperature sensor or the GPU fan, because the temps hit 88, 89 and the fans don't start, when go to the fan control program and change the fan speed to max they go 100% with the interface showing max rpm speeds in both fans (GPU and CPU) but the temps don't get any lower...

 

[Aeacus]

I suspect the problem is somehow connected to the temperature sensor or the GPU fan, because the temps hit 88, 89

What program you use, to monitor GPU temps?

Since it could be either faulty program (trying with 2nd one clears that). Or it could be faulty/dead sensor as well. Though, if it is sensor issue, then it can be quite bad.

In a nutshell, temp sensors within hardware are like nerve ends. If you feel pain, then there's something wrong. If pain is great, issue is severe. But if nervous system doesn't work at all, you won't feel any pain, and thus, you'll never know if something is wrong.

 

[futureal]

What program you use, to monitor GPU temps?

Since it could be either faulty program (trying with 2nd one clears that). Or it could be faulty/dead sensor as well. Though, if it is sensor issue, then it can be quite bad.

In a nutshell, temp sensors within hardware are like nerve ends. If you feel pain, then there's something wrong. If pain is great, issue is severe. But if nervous system doesn't work at all, you won't feel any pain, and thus, you'll never know if something is wrong.

I check the temps in different software like throttlestop, the Gigabyte software...all show high temps and I "felt" it in games because the clock slows and the fps drop...
In other post someone said it may be related to the thermal pads, because I used a 1mm then a 1.5mm and looks like the chassis model for this laptop in some places uses 0.5mm
I asked the gigabyte support if they can tell the thickness of the thermal pads that they use! But yeah is it is the sensor is tough

 

[Aeacus]

What HWinfo64 says? Since that's the most reliable PC monitoring tool, with also loads of sensors shown in there.
Link: https://www.hwinfo.com/download/

Though, GPU boosting into 80+C can be normal.
From another topic:

You can set the temperature limit lower if you want, but the GPUs are designed to boost up to 83C, up to the power and voltage limits set. You could also increase the fan speed, the card may boost higher and stay at 83C or it will reach its power limits and run cooler.

Source: https://forums.tomshardware.com/threads/asus-tuf-3060ti-83°-max-load.3783427/#post-22841651

And when you're using a bit thicker thermal pads that usual, rather than GPU setting at ~83C, it sets 86-87C.

 

[Eximo]

Your different height thermal pads could be causing you to get insufficient contact with the GPU die.

Might invest in some thermal putty like some of the GPU builders use. That way any excess will squish out and allow you do more or less do it without knowing which pads need which thickness.

You could also try some dry fits using various thermal pads thickness, or layers of paper, anything that will leave signs of having been in contact. (You can even put a little dab of thermal compound to see if it gets disturbed)