Question Is My 3-week old EVGA 3080 XC3 Ultra dead?

[jimrawr]

So this morning while I was in a game, my PC just randomly turned off. Everything went black. My first thought was the CPU overheating (have AIO, thought maybe pump isn't working). When I tried to turn it back on, nothing would happen. When I press the power button I could hear a really quick flick, and fans would move a few millimeters before shutting down.

So I started to check and make sure everything was plugged in ok, and then I tested the PC without the GPU connected. When I run it without the GPU, the PC will power on just fine. I can't get to windows since the CPU doesn't have onboard graphics, but all the fans are spinning, the motherboard seems to be working just fine, etc. Luckily I have another gaming PC handy, and I tried the 3080 on that PC. Same exact issue. When I turn on the power, it immediately shuts down. When I remove the GPU, that PC works totally fine.

I tried one more thing. If the GPU has only the RIGHT SIDE 8 pin power connected, the PC will turn on. As soon as I plug in the LEFT SIDE 8 pin, I have the problem. Also, there are red LEDS above the left side pins. GPU was being powered by separate cables, no daisy-chaining.

So I am pretty certain the GPU is dead, but wondering if anyone had a similar experience and if there is anything that can be done besides RMA.

 

[Muadib]

NM, I see that yours was fed by two separate cables. Do you have another PSU that you can try? What psu are you using?

 

[jimrawr]

NM, I see that yours was fed by two separate cables. Do you have another PSU that you can try? What psu are you using?

I should have clarified, but for the 2nd PC it's completely different components. Both PCs have a Corsair 750w PSU.

 

[Muadib]

I should have clarified, but for the 2nd PC it's completely different components. Both PCs have a Corsair 750w PSU.

Sorry, but I'm still confused. Are you saying that you tried it in two different computers? If that's a yes, then I agree that it's broken.

 

[jimrawr]

Yes. I have a completely separate gaming rig. I tried the GPU in that PC and got the same results(it works normally without the 3080). PC would quickly turn off completely and nothing powers on.

Two separate systems, no components shared with the same results.

 

[Muadib]

Yes. I have a completely separate gaming rig. I tried the GPU in that PC and got the same results(it works normally without the 3080). PC would quickly turn off completely and nothing powers on.

Two separate systems, no components shared with the same results.

If there's any consolation, Evga will take care of you. Good luck!

 

[GodisanAtheist]

Yes. I have a completely separate gaming rig. I tried the GPU in that PC and got the same results(it works normally without the 3080). PC would quickly turn off completely and nothing powers on.

Two separate systems, no components shared with the same results.

- I guess I'll just cap the thread off by spelling it out then... Yeah, your card is dead.

 

[DAPUNISHER]

Your card has dual bios, did you try the other one?

 

[maluckey1]

Just read this and.......think that something was going to give with a 650w PSU. 650W is borderline with a 3080, and a PSU at its limits is never a good thing,

EVGA may not care (or even check) what killed the GPU for the first RMA. It'd be a shame to lose $$$$$ over a 100-150 dollar PSU if the new card dies.

M

 

[DAPUNISHER]

Just read this and.......think that something was going to give with a 650w PSU. 650W is borderline with a 3080, and a PSU at its limits is never a good thing,

EVGA may not care (or even check) what killed the GPU for the first RMA. It'd be a shame to lose $$$$$ over a 100-150 dollar PSU if the new card dies.

M

They are both 750W. The first PSU did not fail catastrophically either. Making it possible but not probable that it damaged the card. And while his system could be using over 500W under sustained gaming load, it is still not dangerously pushing a quality 750W bronze or better. The 2nd system displaying the same issue, also points to the card being the problem.

I suggested trying the other bios if it has not already been done, because it is on the very short list of other causes. I.E. a corrupted bios. If it works with the other, OP would have an answer.

And don't get me wrong, I like your advice about investing in a PSU that can handle the load while staying in a better efficiency range. Especially given how much that card sells for at the moment. I am simply doubtful the PSU is the source of the failure. There is a thread in this forum about ABS having a bad batch of 3060s. And with all of the issues we saw with the 20 series early on, it would not surprise me if the card is just a lemon.

 

[Leeea]

RMA it!

 

[maluckey1]

They are both 750W. The first PSU did not fail catastrophically either. Making it possible but not probable that it damaged the card. And while his system could be using over 500W under sustained gaming load, it is still not dangerously pushing a quality 750W bronze or better. The 2nd system displaying the same issue, also points to the card being the problem.

I suggested trying the other bios if it has not already been done, because it is on the very short list of other causes. I.E. a corrupted bios. If it works with the other, OP would have an answer.

And don't get me wrong, I like your advice about investing in a PSU that can handle the load while staying in a better efficiency range. Especially given how much that card sells for at the moment. I am simply doubtful the PSU is the source of the failure. There is a thread in this forum about ABS having a bad batch of 3060s. And with all of the issues we saw with the 20 series early on, it would not surprise me if the card is just a lemon.

And you may be onto something. I tinker al lot....(I get bored) and my recently stable setup went black-screen (not uncommon when chasing rankings in benchmarks), so I immediately thought of the OP. I have a 3070, not a 3080 though. I did what the OP couldn't do....booted using the internal graphics (10700k). The video card wasn't recognized, which is odd.

Odd meant "interesting" to me. Cut to the chase= I reset BIOS, rolled back the last MS update and Nvidia driver, changed to 2 (two) separate power cables (instead of one that was daisy chained).

Next I changed the monitor back to the 3070 and restarted, but black screen again. I was paranoid (4 pints into drinking that time), so re-flashed the BIOS..... and it worked. The 3070 is fine and benches are unchanged. Not sure which solved the original issue,, and my buddies couldn't wither (pretty hardcore with LNo2 cooling). I did look at a few other forums, and they mentioned that the GPU should use two separate power cables for the 3070 and up. I should have known better, and I MAY have corrupted something (BIOS or OS), but managed to fix it.

Parts that matter for this post:

EVGA RTX 3070
Corsair Vengeance LPX RAM @ 4133 mhz (18-21-21-40-62)
PSU=Corsair RMX 850
CPU= Intel 10700k @ 5.0Ghz
Motherboard=Gigabyte Aorus Pro AX

M

 

[maluckey1]

Had to edit drunk spelling, and also wanted to agree that if the OP's card still can't be recognized with 2 individual power cables., that it might be RMA time.

 

[aigomorla]

please let us know how long the RMA takes with eVGA especially with GPU shortages.

 

[eek2121]

Note that I have had that happen with a shared PSU cable. In my case it was a complete accident. Switching to 3 independent PSU cables solved my issue for my 3090.

 

[maluckey1]

Also,

As of yesterday, mine black-screened (yet again) after a power-surge (massive thunderstorm where power went out twice). It seems related to a vendor GPU BIOS re-flash to enable resize BAR.

I switched to onboard graphics and disabled CMA support . I then shut down the computer, plugged in the RTX 3070 and rebooted.... the graphics card worked fine. Not sure if this is something fixable by your average setup without onboard graphics?

M

 

[SgtGanja]

So i hope i can be your savior in this moment, (i could be wrong mind you)

I experienced this problem with my 750W PSU, the 3080 kept triggering some safety mechanism, an upgrade to 1000W PSU fixed all my problems, except for one, I mine Ethereum and as it turns out all KR model 3080's (the ones that are non LHR) all have shotty thermal pads, many people corrected their issues by "rethermaling" all of the GPU and RAM pads. This issue with mine was the EVGA RTX 3080 FTW3 Ultra, but as i understand most KR models have similar issues. The PSU issue basically is across the board, 1000W for 3080's is the untold minimum standard.