2080 Ti core clock problem, please help

[xo4ox]

Here are my system specs:
Intel Core i7 9700k
Gigabyte Z390 Aorus Master
HyperX predator 16gb DDR4 ram
Inno3d 2080 Ti X2 OC
Corsair HX 1200watt power supply
970 evo plus m.2 ssd

I got this system on February of 2019 and been using it since with no issues, the system is water cooled for over 5 months now and had no problems so far. I updated my 2080 ti drivers last sunday and I noticed some stuttering in all my games and after monitoring the GPU with GPUZ, I noticed the core clocks were fluctuating a lot in matter of seconds, goes up to 1900mhz and drops as far as 1350mhz then back up again the whole time. I will attach 2 log files recording this.
Its not an occasional thing, it keeps doing it the whole gaming session and its very annoying.

The major fixes I tried:
1- I reinstalled windows and installed the driver fresh again (problem persists)
2- Downloaded the 442.50 driver version to test if it was a driver issue (problem persists)
3- Found the exact match of my GPUs bios on techpowerup and flashed it thinking it might be a corrupted bios (problem persists)

I ran out of ideas to fix this issue and for the life in me I cant find anyone with this problem, the only thing i can think of that i did wrong was when i updated the drivers the first time before reinstalling windows I left MSI afterburner running with an overclock set on the GPU, forgot to reset it and stop it from running at startup, thats why i considered reflashing the bios since i thought I corrupted it somehow with that.

(find the attached texts for the logs)

Update: It was a temperature problem. I removed the block again and reapplied thermal paste on the core and replaced the thermal pads on the vrms and memory. its back to normal. must have been a mounting issue since temps dropped about 20c. I feel silly for not checking better before but its working now!

 

[the54thvoid]

If you were using Afterburner on a profile, it implies you overclocked the GPU. What OC settings did you use. There are reports that memory OC have resulted in cards failing, though the incidence of it happening is within normal percentiles.

Also, Afterburner wouldn't mess with the card so much - all you'd have to do is reset the card via AB and apply OC again.

Edit: You say the system is watercooled - is that a 3rd party wateblock on the gpu? If so, that's the first culprit - thermal paste, and block fitment could be the problem.

 

[xo4ox]

If you were using Afterburner on a profile, it implies you overclocked the GPU. What OC settings did you use. There are reports that memory OC have resulted in cards failing, though the incidence of it happening is within normal percentiles.

Also, Afterburner wouldn't mess with the card so much - all you'd have to do is reset the card via AB and apply OC again.

Edit: You say the system is watercooled - is that a 3rd party wateblock on the gpu? If so, that's the first culprit - thermal paste, and block fitment could be the problem.

It was a modest overclock, the profile was +100 on the core and +500 on the ram. And i did a reset and cleared the profile after that.

Edit: and I also raised the temp limit and the voltage, but as far as I know the voltage one is basically placebo

2nd edit: yes i used EK waterblocks for the cooling. I could check the mount again but the temps didnt change for the last 5 months when i did the watercooling. i can try mounting it again

 

[the54thvoid]

Under a standard Afterburner unstable overclock, I would get a crash to desktop with AB on low clocks. I'd have to reset AB clocks to default and restart. Usually the game would freeze on a static screen. If you're not getting that, it sounds like some form of throttling, which is usually heat or power related. Try the card at default clocks and see what happens. If it does it at default, there are definitely issues.

 

[xo4ox]

There were no graphical glitches or any artifacts present at all, and those logs were on a stock configuration for the gpu, if I try an overclock it will hit 2085mhz instead of 1900mhz but the fluctuation is still present. ill try a remount in a bit and post back, although temps never exceed 70-71c the whole time. Starting to accept that the GPU might have be defective now

 

[the54thvoid]

Could be time for an RMA.

 

[xo4ox]

Update: It was a temperature problem. I removed the block again and reapplied thermal paste on the core and replaced the thermal pads on the vrms and memory. its back to normal. must have been a mounting issue since temps dropped about 20c. I feel silly for not checking better before but its working now!

 

[the54thvoid]

Update: It was a temperature problem. I removed the block again and reapplied thermal paste on the core and replaced the thermal pads on the vrms and memory. its back to normal. must have been a mounting issue since temps dropped about 20c. I feel silly for not checking better before but its working now!

Nice result