Question Is it Possible to Disable Faulty Temperature Sensor?

invinciblewitzi

Commendable
Feb 24, 2018
7
0
1,510
0
Hey everyone,

A while back I started having an issue where the signal to my monitor would suddenly be cut, but I could hear audio continue to play on my computer. After a while, I figured out that a faulty temperature sensor was the issue (it read that my GPU was running at ~95 °C on boot), and used ASUS GPU TweakII to raise the target GPU temperature to 92 °C. This fix worked for a while, but the issue is starting to occur again, and more frequently as well. I know for a fact that my GPU is running nowhere near 90 °C (or even 60 °C) because right as the signal is cut due to the automatic shutdown of my GPU due to a faulty high-temperature reading, I placed my hand on the GPU, and it was warm, but definitely not anywhere as hot as 90 °C. Every other temperature sensor in my system is fine and reads about the same values, but it's the GPU temperature sensor that is way off. The cooling in my system is great overall, I don't think I've ever gone above 60 °C on any of my components, so clearly the faulty temperature sensor is the issue.
Some more information: it's only the GPU that gets automatically shut down, the signal to my monitor may be lost, but I can still talk to people on Skype or hear music that I was playing. Even online games continue to run, as I continue to hear sound effects and music coming from them.
At this point, the best fix would be to disable the faulty temperature sensor or to disable the automatic shutdown of the GPU because of the faulty sensor's high-temperature reading. Is either of these solutions possible?
Thanks in advance.

PC Specs:
CPU -- Intel i9-9900k
GPU -- EVGA GeForce GTX 1080 FTW
RAM -- Corsair Vengeance 4x8 GB sticks
Storage -- Two Samsung SSD 970 EVO, 1TB, and 500 GB, both M.2
One Samsung SSD 750 EVO, 500 GB, SATA
Motherboard -- Gigabyte Z390 Aorus Elite
PSU -- Corsair RM750
 
First thing first : How can you be absolute sure you have a false reading?

How long since you bought the GPU? Have there being a problem for long time? Did you made any changes to the system before the first occurrence? When does it occur - ingame only?

Info : Holding your hand on the surface of the cooler doesn't prove false reading. Bad thermal contact may explain the issue - i.e. must probably apply new thermal paste, especially if you use something else than stock cooler.
 

invinciblewitzi

Commendable
Feb 24, 2018
7
0
1,510
0
First thing first : How can you be absolute sure you have a false reading?

How long since you bought the GPU? Have there being a problem for long time? Did you made any changes to the system before the first occurrence? When does it occur - ingame only?

Info : Holding your hand on the surface of the cooler doesn't prove false reading. Bad thermal contact may explain the issue - i.e. must probably apply new thermal paste, especially if you use something else than stock cooler.
I got my GPU in 2016. I'm thinking you're right in that bad thermal contact is the issue, as I did not have this issue until I moved last month, so I'll try applying new thermal paste. I did remove the stock cooler because the fans on my GPU stopped spinning shortly before I moved (somehow they got disconnected), which would explain the timing. Thank you for the suggestion :)
 

invinciblewitzi

Commendable
Feb 24, 2018
7
0
1,510
0
First thing first : How can you be absolute sure you have a false reading?

How long since you bought the GPU? Have there being a problem for long time? Did you made any changes to the system before the first occurrence? When does it occur - ingame only?

Info : Holding your hand on the surface of the cooler doesn't prove false reading. Bad thermal contact may explain the issue - i.e. must probably apply new thermal paste, especially if you use something else than stock cooler.
So I finally got hold of some thermal paste and applied it. Temperatures are the same. I decided to let the GPU and memory clocks run as high as possible for a period of time (~20 minutes), and the temperatures are the same as when I allow the clock to fluctuate depending on the temperature. Also, something to note: I let my computer run for a while, shut it off, removed the cooler, and was able to handle the graphics card perfectly fine, it was not hot at all (it took me less than 5 minutes to do all this). If anything, it was actually kind of cold. So really, I think my GPU's temperature sensor got screwed up somehow. Below is an image of my GPU temps from startup (the computer hadn't been running for over 7 hours) and as you can see, the temp sensor instantly reads about 93 °C. I think my best bet will be to buy a new GPU, seeing that my warranty has already expired :/

 

ASK THE COMMUNITY