MSI 2080 Ti Gaming X Trio | Hotspot 105 Degrees!

[FlintSH]

Hey everybody. I've been having some serious issues with my GPU recently with high temps.

I noticed this first a couple weeks ago when I was setting a fan curve in Afterburner, I realized that my fan curve was being overridden.
GPU temps were fine, max like 80c under load, but then I looked at the hot spot.

The GPU's hot spot temperature is hitting 100-107c under load. I have observed that over 105c is when the fans take over and ramp up to around 3000RPM. 
I don't really know how long this has been happening, as when gaming I have headphones on that are great at cancelling out noise, but now I've noticed and have been really concerned.

Things I've Isolated/Tried:
1. Checked HWiNFO for performance limiters (screenshot). 
2. Made sure my airflow is good (it is).

3. Disabled my overclock to see if it fixes (it doesn't).
4. Checked GPU memory temps too to see if they correlate (screenshot).
5. Reinstalled graphics drivers.
6. Updated BIOS.

All GPU Sensor Data:

Spoiler

 

I've tried a lot and I'm really not sure what to do at this point. I'd also like to mention that while HWiNFO says I am not thermal throttling, I have noticed decreased performance on titles where this card should be crushing FPS.
Would appreciate some help, advice, or anything really. Quite lost! Thanks.

[Bombastinator]

1 hour ago, FlintSH said:

Hey everybody. I've been having some serious issues with my GPU recently with high temps.

I noticed this first a couple weeks ago when I was setting a fan curve in Afterburner, I realized that my fan curve was being overridden.
GPU temps were fine, max like 80c under load, but then I looked at the hot spot.

The GPU's hot spot temperature is hitting 100-107c under load. I have observed that over 105c is when the fans take over and ramp up to around 3000RPM. 
I don't really know how long this has been happening, as when gaming I have headphones on that are great at cancelling out noise, but now I've noticed and have been really concerned.

Things I've Isolated/Tried:
1. Checked HWiNFO for performance limiters (screenshot). 
2. Made sure my airflow is good (it is).

3. Disabled my overclock to see if it fixes (it doesn't).
4. Checked GPU memory temps too to see if they correlate (screenshot).
5. Reinstalled graphics drivers.
6. Updated BIOS.

All GPU Sensor Data:

  Reveal hidden contents

 

I've tried a lot and I'm really not sure what to do at this point. I'd also like to mention that while HWiNFO says I am not thermal throttling, I have noticed decreased performance on titles where this card should be crushing FPS.
Would appreciate some help, advice, or anything really. Quite lost! Thanks.

Overclock where?  There are multiple systems in a machine that can be overclocked individually.  Also some stuff that doesn’t seem like an OC actually is.   Specs may be useful though they may also be on one of those links.  I’m trying to not open any internet link I’m not dead sure of lately.)  myself I set my fan curves in BIOS but I understand that’s not reliable anymore either because various stuff can override it.

Edited April 30, 2022 by Bombastinator

[FlintSH]

21 hours ago, Bombastinator said:

Overclock where?  There are multiple systems in a machine that can be overclocked individually.  Also some stuff that doesn’t seem like an OC actually is.   Specs may be useful though they may also be on one of those links.  I’m trying to not open any internet link I’m not dead sure of lately.)  myself I set my fan curves in BIOS but I understand that’s not reliable anymore either because various stuff can override it.

I disabled the Overclock on my GPU in afterburner to isolate whether the hot spot temp issues were correlated. They weren't, with and without OC my problem is still persisting.

[Bombastinator]

2 hours ago, FlintSH said:

I disabled the Overclock on my GPU in afterburner to isolate whether the hot spot temp issues were correlated. They weren't, with and without OC my problem is still persisting.

Heat issues are sometimes factors of case layout.  Hotspots IN video cards, which I assume you are differentiating from a more general heat problem, is something I’m not so familiar with. Though. Some modern cards are known for them.

Edited May 1, 2022 by Bombastinator

[ApolloX75]

On 4/29/2022 at 11:10 PM, FlintSH said:

The GPU's hot spot temperature is hitting 100-107c under load. I have observed that over 105c is when the fans take over and ramp up to around 3000RPM. 
I don't really know how long this has been happening, as when gaming I have headphones on that are great at cancelling out noise, but now I've noticed and have been really concerned.

Hot spot temps like that are usually indicative of either bad thermal paste or bad thermal pads. Or both. Can also be poor mounting pressure but honestly that's not the case here since this is a new problem on a previously functioning card.

 

Can also be attributed to dust buildup in the heatsink of the card but less likely on it's own, usually it's just another cause of the symptom in combination with the other two.

 

You're probably going to need to order sheets of 1mm and 2mm thermal pad and some good paste then tear the card down and service it. It's honestly not as hard as it sounds just need some patience, a little iso alcohol, a clean area to work and a good precision driver set.

[FlintSH]

21 hours ago, ApolloX75 said:

Hot spot temps like that are usually indicative of either bad thermal paste or bad thermal pads. Or both. Can also be poor mounting pressure but honestly that's not the case here since this is a new problem on a previously functioning card.

 

Can also be attributed to dust buildup in the heatsink of the card but less likely on it's own, usually it's just another cause of the symptom in combination with the other two.

 

You're probably going to need to order sheets of 1mm and 2mm thermal pad and some good paste then tear the card down and service it. It's honestly not as hard as it sounds just need some patience, a little iso alcohol, a clean area to work and a good precision driver set.

Yeah. I've been really hoping I wasn't going to have to service the card due to its high value but I guess I don't really have a choice here.

[ApolloX75]

28 minutes ago, FlintSH said:

Yeah. I've been really hoping I wasn't going to have to service the card due to its high value but I guess I don't really have a choice here.

I've done it on probably a dozen high end cards over the last couple years now. Everything from just cleaning and repasting a simple 980 Ti to installing water blocks on my Zotac 2080 Ti and Gigabyte 3080. It's honestly not nearly as difficult as it looks, and once you do it the first time you'll wonder why you were worried in the first place.

 

Like I said just keep your patience in check, don't rush, one screw at a time and don't force anything. Once you get all the screws out of the card frame holding the heatsink down, it will be stiff and it will be difficult to separate because the paste is hardened and the pads are sticky. Just keep your cool and gently work it apart.

[EmileRoss0102]

Hi All, I had the same issue with my MSI RTX 2080. GPU @ 83C and GPU Hot Spot at 107 C with fans almost 100% with MSI kombustor stress test and under hectic game loads as well. 

I took the card apart and applied liquid metal (the safe way).

 

With the same fan speed the GPU was 48 C and Hot Spot 59 C.

 

did the same test with auto fan speed (40%ish) and GPU was 63 C and hot spot 74 C so this is what I consider successful. I could feel the heat sink and case hotter due to the increase in thermal transfer. I’ll show some steps if you guys want to see.

[Sumtinwon]

I see this is an older post. sorry for late reply.  had the exact same issues.   so i tought no my €1500 card aint broken right? light panic.   so i searched for new thermal pads ( yes the exact same thickness and betther quality). replaced them..... nothing.... Resetted the cards bios to when i bought it. rest drivers. you know the drill...

 

50*c idle in win 11

playing ovewrwatch instantly ramps up to 80-83.   still workfs fine tho

started to play with airflow inside my case could handle the heat.  ( 8 casefans so no overheating) after many online searched i always came out to 1 point. its a manufacturing fault of something. cards works perfectly but get instant high uncontrollable temps.

so i changed to a 4070 ( my cpu is a AM4 r9 3900x)  so 4070 was the max i could install to not have or minimal bottleneck.

with the new themal pads i decreased temps with what i believed it was 1.26celcius so that was the blidn test to check if they where right or wrong. and it was good. so yea

 

with the 4070 my temps are verry normal idle 35-42   load max i saw was 66*c.

 

I still have the 2080ti and i miss it tbf  but i went for the safe way.

if there is another way to solve this prob . pls trow it in here even after a long time.