GPU voltage spikes

[Mcwrah]

Hey guys so i got this problem... I recently did some cable management in my PC. After starting it up i went to play a game. My voltage kept jumping from 0.980V to 1.063V, and shortly after my pc just shut down and refused to boot. The only way to boot it was to turn off psu and turn it back on. After i did that and pressed the power button, it turned on for 5sec and then immediatelly turned off, and then instantly back on and booted. Now to play any game i have to set max power limit to 120% or the voltage starts to spike and go down and crash my PC again. I have tried different PCIE slot and GPU, but i get same results (spike in voltage and then crash, unless i set my power limit to 120% in msi afterburner). I even tried reinstalling the nvidia drivers but it didnt help. I didnt have this problem before, it suddenly started today.


Its a new system, a new 1080ti AORUS xtreme (1080 asus advanced edition is the backup card that has 0 problems). The cpu is 2700x + asus crosshair VII. PSU is from old pc which i kept: evga supernova G3 750, which is only 8 months old.
Do you guys have any idea what the problem is? CPU? MOBO or the PSU?

edit: forgot decimals at the voltage, my bad. fixed.

 

[RobCrezz]

The problem sounds like its with the GPU, I would return it for a replacement.

The voltage fluctuation isnt unusual, but the crashing and instability is.

 

[Shektron]

You must be reading something wrong, because 980 V will probably set things on fire in your PC. We deal with milli-volts in PC, and a maximum of around 12 V, so 980 V is just wrong. I guess you are forgetting the decimal perhaps. GPU core voltages are around the 1-1.2 V mark.

Either way, the issue does sound PSU-related. Can you try running your GPU in a different system?

 

[Mcwrah]

As i mentioned i have 2 gpus(1080 and 1080ti, the 1080 being backup one), it happens with both, in both PCIE slots, also as of now, the fluctiation is suddenly gone and i have stable voltage, made me crash again until i bumped power to 120% for 10minutes and played a game which made it all stable and i was able to put it back to 100% with stable voltage.

edit: yeah pretty much, i forgot the decimals lol, fixed it. I dont really have a system that can handle the gpu, only one with a 450W bronze psu and not sure i want to try it in that lol. Also i think its safe to say that its either motherboard or PSU since no matter which gpu of the 2 gpus i plug in and try out, both have spikes for brief time, unless I bump the power limit and use it in a game or something for few minutes after which it returns to stable.

 

[RobCrezz]

As i mentioned i have 2 gpus(1080 and 1080ti, the 1080 being backup one), it happens with both, in both PCIE slots, also as of now, the fluctiation is suddenly gone and i have stable voltage, made me crash again until i bumped power to 120% for 10minutes and played a game which made it all stable and i was able to put it back to 100% with stable voltage.

edit: yeah pretty much, i forgot the decimals lol, fixed it. I dont really have a system that can handle the gpu, only one with a 450W bronze psu and not sure i want to try it in that lol. Also i think its safe to say that its either motherboard or PSU since no matter which gpu of the 2 gpus i plug in and try out, both have spikes for brief time, unless I bump the power limit and use it in a game or something for few minutes after which is return to stable.

Oh sorry, i read it as it all works fine with the spare 1080...

Are you using any cable extentions on your PSU? Can you try one of the spare PCI-E cables and plug it into a different connection on the PSU?

 

[Mcwrah]

As i mentioned i have 2 gpus(1080 and 1080ti, the 1080 being backup one), it happens with both, in both PCIE slots, also as of now, the fluctiation is suddenly gone and i have stable voltage, made me crash again until i bumped power to 120% for 10minutes and played a game which made it all stable and i was able to put it back to 100% with stable voltage.

edit: yeah pretty much, i forgot the decimals lol, fixed it. I dont really have a system that can handle the gpu, only one with a 450W bronze psu and not sure i want to try it in that lol. Also i think its safe to say that its either motherboard or PSU since no matter which gpu of the 2 gpus i plug in and try out, both have spikes for brief time, unless I bump the power limit and use it in a game or something for few minutes after which is return to stable.

Oh sorry, i read it as it all works fine with the spare 1080...

Are you using any cable extentions on your PSU? Can you try one of the spare PCI-E cables and plug it into a different connection on the PSU?

I havent tried different cables yet but i will. I however, have tried going from VGA 1,2 to 3,4 with same cable and i have no luck there. Nothing changed.

edit: i also should mention i have similar coilwhine on both gpus. I remember my 1080ti coilwhine being a lot more quiet than it is when the voltage spikes happen. Now atm its how it used to be, but when its unstable and makes my PC shut down it gets really high pitched and annoying.

Edit2: Tried the cables, still same. I also went into bios to reset it, but before i could do anything my bios froze and after waiting for 2min i had to force shutdown my PC. what the hell is going on lol.

 

[RobCrezz]

Thats really weird issue.

You have a really good PSU, so I was doubting it was that, but its possible its just faulty.

Or you can try going back to basics, running the bare minimum to rule out some other hardware.

 

[Mcwrah]

Hmm yeah, i might go have it replaced (where i bought it from, theres an extra payment which is basically 10% price of the item added up, and if you have any problems with it they will replace it without waiting) and i have that so i could go and have it replaced and see if the problem persists. If it does well it might be the psu because the only other things i could unplug are kraken x62, which up until i did the cable management worked just fine, and hell it even works fine, i dont usually get above 44C in games with it (push/pull config). For example, FFXIV which i played while this started, i had and still am playing it (when i put my gpu to 120% power limit so it doesnt fluctuate), i get 38C on CPU, the lights work, pump works too, all the fans are spinning without a problem. The only thing i could really unplug apart from case fans is the 2hdds and 2ssds. Sad that until i did the whole cable management and took out my psu and put it back... everything was fine really. Also i didnt touch anything lol, just took out the cables and made them look neater and less messy.

 

[Mcwrah]

Thats really weird issue.

You have a really good PSU, so I was doubting it was that, but its possible its just faulty.

Or you can try going back to basics, running the bare minimum to rule out some other hardware.

So before having it replaced i did the only thing i really could which is: I took out the psu and put it into my old PC, which is basically i5 4690k and 200€ motherboard from asus and 16gb ddr3 ram. I also took both gpus and tried them in. After going into a game, i had 980mV to 1063mV spikes same as on my main PC, only fixed if i used evga precision X or MSI afterburner and bumped power target to 120%, which made both cards have stable 1050-1063mV voltage(which also fixed fluctuating clocks that went from 1883-1963 and were making my fps fluctuate too from 192-199, with 120% power target it went to 205 on the 1080ti). With both gpus. I also tried to see what 3dMark would say but it just outright blackscreened me and drivers crashed. With both GPUs. Even reinstalled drivers with DDU in safe mode. I think its safe to say my PSU is faulty, so i am going to have it replaced asap.