Question Graphics card died after an interrupted Windows update ?

[wordweaver]

Full description:

I've had the problem, that after the screen goes dark (power saving after 30 minutes) there was no image coming from the gpu when using the mouse for... quite some time now on two different monitors (TV, curved Samsung) but a restart always solved the problem (replugging the cable did seldomly work, no matter if hdmi or dp)

so some weeks ago (day before my vacation) there was a windows update that got stuck. after 2 hours of being stuck, I pressed the reset button... and the PC got stuck in a starting loop without even showing bios.

my 3080fe is water-cooled... so it's rather annoying to check the graphics card. but after having no other ideas I unplugged it and switched to integrated graphics. integrated graphics worked without any problems.

once or twice I was able to start the system with the graphics card plugged but the screen being plugged into the integrated graphics port: no graphics card was shown whatsoever in Hardware, neither the system.or nvida could find the installed graphics card.

so I really don't know if the graphics card died just outside of warranty or if there's another problem, connected to the failed update? and this being a custom made Waterloop... any ideas of checking for the source of the problem?

like: it could be the pcie AFAIK (graphics was plugged in with an adapter, problem.continues without using the adapter, so the slot was never taking the full weight of the graphics card, but also directly plugging it in did not work, the adapter is not the problem)

It could be the power supply... but the system worked for years on end, and while starting it does not use full power and I had no crashed while gaming, so power should be fine)

I tried all rams separately and in different slots, they were fine too.

so it seems, that quite probably it's the gpu (the most expensive part 🙁 ) but how can I be sure that it was not the motherboard killing the gpu? 😅

any tips and suggestions are welcome, like did I forget to check anything (I order a new psu and gpunalready before coming home like... with amazon 30days return it's nice. I would first switch the psu (that's already a lot of recabling 😅) and if that would njot help I'd switch to an 4080... but still, I can't really check if the graphics card itself is faulty, as it is water-cooled..)

thank you already for your input!

 

[Lutfij]

Welcome to the forums, newcomer!

When posting a thread of troubleshooting nature, it's customary to include your full system's specs. Please list the specs to your build like so:
CPU:
CPU cooler:
Motherboard:
Ram:
SSD/HDD:
GPU:
PSU:
Chassis:
OS:
Monitor:
include the age of the PSU apart from it's make and model.

If you think it's the discrete GPU, remove it and try working with the iGPU. If you don't have an iGPU, you will need another discrete GPU to replace the one you already have. If you're working with a PCIe riser cable, remove that and work with the discrete natively on the motherboard's slot.

 

[wordweaver]

I did try it without the riser cable, same problem
integrated graphics do work

cpu: i9700k
cooler: watercoolerblock, alphacool

as I'm not at home I can't supply the rest of the specs from the top of my head, but it worked the last three years without any problems

 

[Aeacus]

any ideas of checking for the source of the problem?

PSU make and model (or part number) is? Also, how old the PSU is, and was the PSU bought new or used/refurbished?

Btw, same question to your new PSU as well.

so I really don't know if the graphics card died just outside of warranty or if there's another problem, connected to the failed update?

Well, your RTX 3080 is toast. Also, i don't think it died with failed update but earlier. Failed update was just the last nail in the coffin.

that after the screen goes dark (power saving after 30 minutes) there was no image coming from the gpu when using the mouse for... quite some time now on two different monitors (TV, curved Samsung) but a restart always solved the problem

A sign of failing GPU. You should've dealt with it back then.

but still, I can't really check if the graphics card itself is faulty, as it is water-cooled..)

I'm wondering; why use open-loop if:
* It is so high maintenance (flush the system every 3 months + deep clean once a year)
* It makes troubleshooting (hardware replacement) tedious?

Pretty looks and a bit lower temps, IMO, isn't enough of a reason to run open-loop.

 

[wordweaver]

That's what I could find fast

but nice to know that it quite probably is the graphics card. I did the upgrade during covid, like aircooling was so loud, that it was annoying while videoconferencing (and gaming at the same time of 😅) and just for the fun of it (I like to build things) and I have very little problems with the maintenance, no buildup inside whatsoever, so it's more like cleaning everything once a year

30/10/2020 Nvidia 3080 FE

new psu will be 1000w rmx shift corsair
new gpunrtx 4080 kfa2, this time without watercooling so just the cpubwill be water-cooled


17/10/2019

1x Netzteil CORSAIR 750W RM750 ATX Modular (80+Gold)

1x Kühler Thermalright Le Grand Macho RT

3x Fractal D. Prisma SL-14 rd | FD-FAN-PRI-SL14-RD

1x Intel Core i7-9700K R0 3,6 GHz (Coffee Lake) Sockel 1151 - tray

1x Fractal D. Define R6 C Blackout TG bk

 

[Aeacus]

Corsair RM750 is good quality PSU and i doubt it was the reason behind GPU dying and eventually failing. Though, with RTX 4080, 1kW PSU is suggested regardless.
RM1000x Shift is also good PSU and it's also ATX 3.0. But due to it's unique design, you may have little room to route the cables from PSU.

Regarding GPU, it is difficult to say what made it die on you. There are several aspects and any one could've been culprit. Maybe even all combined or none of them. Those are: custom cooling (water block), riser cable, GPU OC, Nvidia PCB.

 

[wordweaver]

Corsair RM750 is good quality PSU and i doubt it was the reason behind GPU dying and eventually failing. Though, with RTX 4080, 1kW PSU is suggested regardless.
RM1000x Shift is also good PSU and it's also ATX 3.0. But due to it's unique design, you may have little room to route the cables from PSU.

Regarding GPU, it is difficult to say what made it die on you. There are several aspects and any one could've been culprit. Maybe even all combined or none of them. Those are: custom cooling (water block), riser cable, GPU OC, Nvidia PCB.

Well, this time it's going to be plugged in directly with a riser stand and without a riser cable adapter and it stays air cooled (the notice culprit was the cpu fan after all, so the cpu will stay water-cooled as long as the pump works and the setup will be migrated to upgrades systems through time as I really like the optics of it - the pc is mounted behind my TV usually, so looks do matter a bit 😀)

and I can exclude oc person as I never oc anything 😀