Question ASUS GeForce RTX 2080 Ti 11GB DUAL OC Thermal Pads Replacement Thickness

Flourix

Commendable
Jan 4, 2020
9
0
1,510
Hello!

My graphics card ASUS GeForce RTX 2080 Ti 11GB DUAL OC was bought October 2018.

Recently the temperatures are going quite high and I need to keep a higher fan curve which is loud, both on chassi fans and GPU fans. At a aggressive fan curve temps are around 80-85c in demanding games on high graphic settings, like big MMOs or good graphical open world games. But the difference between GPU temp and Hotspot temp varies about +15-20c making hotspot reach 100+ Celsius most of the time. It got me a bit worried since I read optimal is a difference of +5-10c and also with hotspot above 100c do I need to be worried for overheating?

I recently ordered new thermal paste and replaced old original one which didn't cover some spots. But when I open up I realised I forgot completely about the thermal pads and possible to exchange them to. They didn't look too bad condition as some others I've seen in videos - but I guess new ones from 2022 are better performing than original ones anyways .

My temperature with the new thermal paste is about the same and it didn't really change much. I would like to not open up graphic card an extra time now just to measure the pad thickness before I order and mess upp the thermal paste. So I thought I'd ask here if anyone of you have the correct thickness for me to go for? Or know where I can find one of those drawings that show every pad placement and thickness used like I've seen for other cards? Didn't find one when googling for my graphic card sadly.

If you have any info I would be very grateful and then I order and try and replace the pads and cleanup and reapply new paste again I suppose since I probably would mess it up while opening it up again.

Thanks for all the help!
-Robin
 
How did you apply the new thermal paste on the gpu? Did you spread it all over the entire surface? When I changed my thermal paste (3080) my gpu temps dropped 15c. The pads are really for the vram and vrm cooling. I left the same ones in mine, but I know other people have found melted or disintegrated pads in theirs. What is your room temperature? Is your case well vented with fans? Here is a chart for placement and thickness of new pads. https://prnt.sc/ZQtFX3HuHtXI
 
Yes I used a trimmed down plastic knife and spread it evenly, maybe it was a bit thick layer though because I think I put a bit too much first and removed some, but evenly spread. I read that for GPU it's better to spread evenly yourself instead of doing the CPU method of just putting few dots on.

Edit: I used Artic silver 5 Thermal paste now. You recommend something else?

My case is a Fractal Design Define R6 with tempered glass side. And chassi fans are some powerful Noctua fans that I want to keep around maximum 1000rpm if possible for reduced sound. Room temperature is now in the summer around 20-26c and in the winter like maximum 20 but usually I don't heat up as much and tolerate living in 15-20c.

My pads was not melted or completely dried out, but looked like they were pressed down to a smaller thickness than original unevenly on their surface so to speak.

But hotspot is only for the GPU then? And pads won't help you think? Only for memory temps?
 
But the difference between GPU temp and Hotspot temp varies about +15-20c making hotspot reach 100+ Celsius most of the time. It got me a bit worried since I read optimal is a difference of +5-10c and also with hotspot above 100c do I need to be worried for overheating?
The former is normal - more like 10-20C; my 1080Ti has carried a gap of 14C between 3 different coolers.
The latter, I don't think is possible unless it's under LN2...


Unfortunately, replacing thermal pads is not that simple - well actually, the simplest method is that someone has already successfully replaced the pads, and left a post or article on the EXACT brand, model, and size of pads used for your specific card.
If you can't get hold of such info, you are in for a ride, and starting in hard mode.
It is not enough to know pad thickness, as they come in different degrees of hardness, even across different brands. If the pads are too hard, you end up with worse core and hotspot, as well as risk breaking the PCB.
If the pads are too soft, then the Vram stays hot, because it's not making firm enough contact with the heatsink. No risk of physical damage though.
Do not use the W/mk rating to determine how good thermal transfer is. On the contrary, use it to determine how hard the pads are; they have to be for those higher ratings. The pads the manufacturers use tends to be on the softer side.
 
Sounds like you spread it correctly on the gpu. And the Artic is fine to use. Also my gpu only dropped 10c not 15c with new thermal paste. You said you have an aggressive fan curve set, could you post a pic of it? I wouldn't worry about hotspot temps, that is just telling you the hottest spot detected by the sensors. If the gpu isn't shutting down, and the temps stay controlled you are ok. I would try maybe turning case fans up to move heated air out. And let's see what fan curve ideas we could come up with after looking at yours.
 
Ok, maybe I can then make a little bit more gently curves if hotspot temp is nothing to worry about if it reaches 100+. And replacing the Pads sounded a bit complicated to get right now when you described it like that.

One thing I noticed is hotspot temp. Is following vRam temp. Quite similar. So maybe hotspot temp is also taking memory temps in. I'm using Argus Monitor to watch my temps.

Here is my fan curves. Chassi fans at 20% standard is around 700rpm. And then I have a % of speed that is OK Sound wise with headset. If possible I would run them a bit more silent though if hotspot temp is nothing to worry about.

GPU fan curve
https://1drv.ms/u/s!AmeOm749yq5sywJpq-IdTr-pqvIT

Chassi fans curve
https://1drv.ms/u/s!AmeOm749yq5sywEAGiQB7kRQQYW7
 
Sounds like you spread it correctly on the gpu. And the Artic is fine to use. Also my gpu only dropped 10c not 15c with new thermal paste. You said you have an aggressive fan curve set, could you post a pic of it? I wouldn't worry about hotspot temps, that is just telling you the hottest spot detected by the sensors. If the gpu isn't shutting down, and the temps stay controlled you are ok. I would try maybe turning case fans up to move heated air out. And let's see what fan curve ideas we could come up with after looking at yours.

Do you want or need some more information or pictures of the curves? Just tell me and I'll look into it.