RTX 3080 indicates permanent 60°C in task manager

[The Grin]

This happens since this morning. Is this a bug ?
The fans are turning even on idle but the general power plan is "balanced".
Usually I have 30° ?
Even while gmaing it's at 60° :s

[_I_]

how is the gpu load at idle?
check with gpuz

[The Grin]

_I_ původně napsal:

how is the gpu load at idle?
check with gpuz

So, I just instaleld it.
The temperature indicated is 61°C (with only steam running).
There is something really wrong.
The tempered glass is somewhat hot but I am far from being able to cook an egg on it.

Like someone esle stated maybe I will have to change the heating pads.

PS: fun fact => during gaming, the temperature stays at 61°C !!!
Maybe my GPU thinks it's a lucky number and draws it just like that :s

[GOD RAYS ON ULTRA™]

Task Manager isn't the utility you should be using to judge metrics. You should be using gpuz.

[_I_]

maybe the fan curves are set to keep it around that temp

at higher load it may have higher fan speeds and keep the temps around the same mark

but keep gpuz running while gaming, it logs min/max temps fan speeds and loads

[ˢᵈˣ FatCat]

gpuz is still a software, they just take a data , from a sensor, a sensor inside your motherboard/gpu/cpu
still a software
gpuz is just more detailed, numbers we , casual dont care, wont care, so i choose the things have same result, but more fast,more convinient

[Bad 💀 Motha]

Yes make the effort to take any PC tower and remove any panels and filters and use an air blower and blow out all dust from entire system and clean the filters if present. Do this outside if possible with regards to blowing out all the dust. Every 3-6 months or as often as needed based on build up on the fans and such

[[☥] - CJ -]

My guess would be if you had Prefer Max Performance set in the NVIDIA Control Panel that something could be keeping the clocks raised.

You could always add the GPU tab to the Task Manager Process list to see if anything is using the GPU when it shouldnt be.

Otherwise i would just install the latest driver if you havent done so yet, i seem to recall a temperature being stuck bug or something like that in a previous driver

[Bad 💀 Motha]

☥ - CJ - původně napsal:

My guess would be if you had Prefer Max Performance set in the NVIDIA Control Panel that something could be keeping the clocks raised.

You could always add the GPU tab to the Task Manager Process list to see if anything is using the GPU when it shouldnt be.

Otherwise i would just install the latest driver if you havent done so yet, i seem to recall a temperature being stuck bug or something like that in a previous driver

I do that on every PC to ever have an NVIDIA GPU in it; they still should be around 30-40*C at most even when the 3D Clocks are being pushed 24/7

I suspect the user has a bugged GPU Driver or maybe even a bugged Windows Update causing such behavior from the GPU. Wipe that all out (everything NVIDIA) using the DDU app, run it and change the options to allow for safe mode, then close the options and exit the DDU app, then relaunch it. Now it should come up and ask how to run it, select Safe Mode + Restart. Then WinOS should reboot into safe mode and autorun DDU for that single instance. Then under Graphics/GPU select NVIDIA and click "Clean and Restart"

If you were already on the most recent GPU Driver before, try an older one, such as 555.99 or one of the others such as 563.xx

While older drivers have security holes, NVIDIA suggests that their 553.xx and later are much safer. However I found above 555.99 to be rather pointless unless running RTX 40 series.

[_I_]

^ agreed

even at performance/boost clocks and idle, gpu fans off still <50c even if the room is warm and most case fans at low rpm

60c is still no harm, as long as the gpu is not maxing out above 90c and throttling its fine

[Bad 💀 Motha]

It takes having my room reach like 30*C+ to have my GPU such as RTX 3080 or 3080 Ti go above 65-70 and that's on nearly full loads for many hours at a time. And no, they are not liquid-cooled either. Anytime I'm not using a GPU work app or running a game, they run at 30-32*C

[The Grin]

Hello everyone.

I am so glad you are all answering and trying to figure out with me what this problem is .
I was so busy yesterday but I'll try undusting my pc today.
If nothing changes then I'll reisntall the previous driver (I did not have this issue while using the pc with the previous driver!) and see what it gives.

The only thing I don't want to see ( trying to boost positivity):
- fans going haywire on a " Windows desktop" pc with only the mouse moving.
- T° going gradually up from steady 38° (should be 30...like usual on start up) to permanent 60.

Wish me luck today is the day :s

[Skkooomer Lord]

Dial 911 and ask for fire brigade.

[Agenda 2025]

disconnect your wifi router. Windows 11 new version has AI running all the time. Will suck a laptops battery dead quick even if off. Cutting its internet helps...but that AI still runnin all the time

[The Grin]

UPDATE:

It seems the last driver (566.03) was at fault and loved tickleing that Graphics Card.

I reinstalled the previous Nvidia driver from October 3rd (565.90).

The pc is not as "noisy" as before, the fans don't spin anymore, the temperatures went down to 39° (not 30 but anyway).

I'll undust all that little machinery and I hope I'll see myself sipping on an Ice Tea if the temps go down a bit further, but all in all it's allready good as it is.

Thanks to everyone trying to help.

[Phénomènes Mystiques]

If you repaste the GPU you also need to replace all the thermal pads within.