[SOLVED] high temperatures after changing thermal paste on gpu

DarkRaiden112

Prominent
Oct 1, 2020
10
0
510
hi i have a xfx rx 570. i get very high temps on it after i changed the thermal paste. before changing the paste i used to get around 80 degree temperature even after undervolting and with fan speed on 80 percent. so i thought maybe the thermal paste had gone bad. now after i replaced the thermal paste i get 83-86 degrees temperature while playing days gone as it consumes 100 watts of power. this seems very high. i have made sure to re do thermal pastes 2 or 3 times to make sure the heatsink was making correct contact and still im getting same results. could using a better thermal paste improve my temperatures by alot? im using one called hy510
 
Solution
i have used the X method and the pea method aand both give the same results. im starting to think its the thermal paste thats the problem

It is. HY510 is bargain bin stuff, the only time I see it being used is for cheap custom PCs at my local store.

Check this out, they did a test on a laptop which uses the same direct die contact as our GPUs, and they found that the delta between HY510 and run-of-the-mill MX4 is about the same as top tier thermal paste (Kryonaut, KPx, etc) to the worst stuff on desktop.

Personally, I also have 2 small tubs of HY510. Temps are nowhere as good as MX-4, as tested on my i3-550. I use it mainly to test contact patch before applying the paste I'm going to use.

You'd want at least an MX-4...

DarkRaiden112

Prominent
Oct 1, 2020
10
0
510
you might not have placed the right amount on the GPU. try different thermal placement methods. for my GPU i used a good pea size dot on mines. for my cpu i used the X method. search youtube on how to apply it.
i have used the X method and the pea method aand both give the same results. im starting to think its the thermal paste thats the problem
 

iPeekYou

Distinguished
Jul 7, 2014
392
77
18,790
i have used the X method and the pea method aand both give the same results. im starting to think its the thermal paste thats the problem

It is. HY510 is bargain bin stuff, the only time I see it being used is for cheap custom PCs at my local store.

Check this out, they did a test on a laptop which uses the same direct die contact as our GPUs, and they found that the delta between HY510 and run-of-the-mill MX4 is about the same as top tier thermal paste (Kryonaut, KPx, etc) to the worst stuff on desktop.

Personally, I also have 2 small tubs of HY510. Temps are nowhere as good as MX-4, as tested on my i3-550. I use it mainly to test contact patch before applying the paste I'm going to use.

You'd want at least an MX-4 level of performance, also a thicker paste for better longevity. Kryonaut is good, but der8auer also recommends repasting each year if you use Kryo on GPU. They will pump out faster than the thicker pastes.

On GPUs, the thermal paste application method matters less unlike on Ryzens. Just make sure to cover the entire die surface and you're good. Bit excessive is okay if you use thermal paste, not liquid metal. I myself also err on the side of excess when doing GPUs.
 
  • Like
Reactions: DarkRaiden112
Solution