[Help Needed] My card's core & mem clocks are spiking then throttling, then spiking, throttle... up and down every second + no signal & black-screens

[Euphoria]

Hi everyone, I really need some help and advice please.

My system is old but still does what I need it to as I mostly play older games plus I do Not have the money to upgrade, the card is a Gigabyte G1 980Ti. I've been happily playing every night when suddenly tonight my fps is bouncing around like crazy; 150 to 30, 150 again, then 30, 150... I alt-tab out and checked Afterburner to find my core and memory clocks are spiking and throttling, jumping up and down every second or two; 1550MHz to 400MHz, back to 1550, 400 and repeat... Then it black-screened but the system was still on just my monitor kept searching for a connection. It's black or white-screened a few times since when I put it under heavy load.
Temps are fine according to Afterburner, under 50c and the fans are spinning but at low rpm because of the low temps.

 

I've tried everything I can think of over the past 4-hours but I can't fix it and I'm starting to worry my card has died, this would destroy me as gaming is pretty much the only outlet I have right now and I couldn't even buy an old second hand 600-series card right now after having to give up my job.

I've tried so far:

• Under-clocking the card
• Under-volting/power limit
• Using DDU in safe mode to uninstall everything Nvidia
• Installing 4 different drivers after uninstalling with DDU each time
• Tried different games and stress tests
• Removing the card and reseating it
• Thoroughly cleaning the card
• Cleaning the PCI-E slot
• Different PCI-E slot
• Different PSU cables in a different modular socket on the PSU
• Plugging the system into a different wall power socket
• Running off a UPS

The only other things I can think of are - striping the card to reapply TIM and thermal pads but I did that about 2-years ago and temps have always been fine, reinstalling Windows or doing a fresh install on another drive in case Windows is causing something, testing it on a different system and/or PSU but I don't have access to one as this is my only desktop and I recently emigrated so I don't know anyone here I could ask.
As said I really can't lose this card right now, so please any advice or ideas would be hugely appreciated. Thank you.

Edited October 10, 2020 by Euphoria
Changed title to make it easier to understand the issue

[Airdragonz]

You can try a different dp/hdmi port and cable on the card and monitor. If that doesn't work, try running something gpu intensive on a fresh install of windows or linux.

[Euphoria]

30 minutes ago, Airdragonz said:

You can try a different dp/hdmi port and cable on the card and monitor. If that doesn't work, try running something gpu intensive on a fresh install of windows or linux.

Could a bad port or cable cause the clock stability issue?

[Airdragonz]

1 minute ago, Euphoria said:

Could a bad port or cable cause the clock stability issue?

If the signal becomes lost, assuming you only have one monitor, then the clocks could drop.

[Euphoria]

17 minutes ago, Airdragonz said:

If the signal becomes lost, assuming you only have one monitor, then the clocks could drop.

Ah right, yeah the core and mem clock will bounce up and down even if the monitor is still displaying or either monitor and I've tried with only one, but to be sure I just tried a different cable and port with no luck.
I will try a fresh OS install tomorrow as it's 3:00am and my brain is shot.