Question 2080 Ti significantly underperforming

Feb 6, 2025
2
0
10
Sooo.. this is gonna be a long one.

I came across an old 2080 Ti of my friend, who recently upgraded his rig and asked me to look at his old GPU.
When he used it (850W PSU dunno the brand, with a 3600x at that time) it worked as it should for about 2 years or so. Then it started going downhill.
The card clocks to like 1350-1550 MHz and voltages go to 0.720 V, power doesnt go higher than like 150 W.
The card HEAVILY underperforms in benchmarks and games (Steel Nomad ~1500 points instead of expected 3700)
I tried it out in my rig (7800x3d, 5400mhz 32gb ram, msi mpg motherboard, 750W be quiet PSU) and it had the exact same issues.
Fanspeed runs on max (3500 rpm) and the temps don't go over 73°C.
No matter what settings I throw in on MSI Afterburner the card does whatever it wants.
I tried forcing 1.000 V and 1950 MHz clocks and 40% fanspeed, which is stable when clicking checkmark on Afterburner.
As soon as I start a benchmark the card the clocks go down to the same stuff as before (1350mhz, 0,720v, 3500rpm, 73°C) and ignores my Afterburner settings.
Using HWinfo and temp limit has never reached.

Also when I simply reduce the power limit to like 70% off the stock settings, the performance improves significantly over 100% power limit (still overall VERY bad performance).
I've found a bunch of the same posts, all of them at least 3-5 years old and ALL OF THEM with the same card:
ASUS ROG STRIX 2080 Ti
Yet not a single thread I've found has a solution.

I tried out:
undervolting + curve editor
reset windows
try different rigs (mb, ram, cpu all different)
energy saving modes in windows on max performance
DDU + fresh install

If anyone can help to not let that poor 2080 Ti rot in the basement, please help me. :)
 
Last edited:
Welcome to the forums, newcomer!

You didn't mention the specs to all the system's you'd dropped the card into/onto. Make and model of the PSU's used to power the GPU?

You could try and disassemble the card, deep clean it and in the process replace the thermal paste and pads with much higher quality products. Though it could very well be the card just giving up the ghost.

I've found a bunch of the same posts, all of them at least 3-5 years old and ALL OF THEM with the same card:
ASUS ROG STRIX 2080 Ti

FYI there was a known issue with the GTX1000 series and 2000 series under the Strix range whereby if you rectified the sag in the card/cooler, the temps would go up instead of down.

DDU + fresh install
You're advised to install the OS in offline mode(for Windows 10 or 11) then install all drivers in an elevated command, i.e, Right click installer>Run as Administrator while offline. If you're running DDU, you should remove all GPU drivers(Intel, AMD and Nvidia) in Safe Mode.
 
  • Like
Reactions: TShpend
Welcome to the forums, newcomer!

You didn't mention the specs to all the system's you'd dropped the card into/onto. Make and model of the PSU's used to power the GPU?

You could try and disassemble the card, deep clean it and in the process replace the thermal paste and pads with much higher quality products. Though it could very well be the card just giving up the ghost.

I've found a bunch of the same posts, all of them at least 3-5 years old and ALL OF THEM with the same card:
ASUS ROG STRIX 2080 Ti

FYI there was a known issue with the GTX1000 series and 2000 series under the Strix range whereby if you rectified the sag in the card/cooler, the temps would go up instead of down.

DDU + fresh install
You're advised to install the OS in offline mode(for Windows 10 or 11) then install all drivers in an elevated command, i.e, Right click installer>Run as Administrator while offline. If you're running DDU, you should remove all GPU drivers(Intel, AMD and Nvidia) in Safe Mode.
My friends specs were:
CPU: R5 3600x
RAM: 3200 MHz 32GB Corsair Vengeance
PSU: 850 W be quiet! PSU (probably pure power too idk)
Mainboard: some ASUS ROG am4 mainboard
OS: Win 10

Mine:
CPU: 7800x3d
RAM: 5400 MHz 32GB Kingston Fury
PSU: 750W be quiet! Pure Power 11 Gold+
Mainboard: MSI MPG B650 Edge WIFI AM5
OS: Win 11

About the disassemble: I ordered thermal pads and paste and hope they'll come in tomorrow and I wanna try that out too, even if the temps look okay, maybe it's just small parts of the VRAM pads broken and not all or something ? No clue about all of that, I've never changed any pads but I will gladly try it out tomorrow (if Amazon is fast enough 😛)

By redctifying the sag you mean "holding up the card"?
I mean the temps are okay, aren't they? with 72°C temps should not be a problem. Also it is very weird the card is running 3500 RPM while on 72°C.
It's just straight up ignoring my 40% fanspeed as soon as I start a benchmark lol.

Installing the OS in offline mode so it doesn't autoinstall drivers I am guessing? I think I won't try another new install though, since I tried different PCs and tried uninstalling them with DDU already (though while online).

Might it be a hardware issue you think? Just straight up broken parts on the GPU I mean, which would be weird for a 2,5y card.
 
Last edited: