After playing Cyberpunk 2077 for about 15 minutes, my computer crashed. When I tried to reboot, I noticed a single red light on my 3080 FTW3 Ultra which I just bought two weeks ago. I replugged everything in with no help. I've read that this may be a common problem and just wanted to make sure that my issue was counted/recorded. My rig is:
3700x
32 gb ram 3200 cas14
Gigabyte x570
750 watt PSU bronze
1 TB m.2
1 TB 2.5 sad

I know my power supply is on the bare minimum but I also have absolutely no overclock on my system. Also, I've seen the problem with much higher PSU's. I tried to contact technical support but have to wait till Monday apparently.

UPDATE: I went to microcenter this morning to get the card tested. It is definitely the card which is broken and nothing else inside my pc. The manager was nice enough to fill out a voucher for me and will give me a call if they get any 3080s in so I can come swap mine out. Seems better than an RMA since it is potentially faster AND I can get a different brand that is hopefully more reliable.

I think I have to wait till Monday to get help first in order to RMA unfortunately. I have no spare card meaning I can't get any work done till this is fixed. Extremely unfortunate.

My card probably went through 15-20 hours max. This was mainly Warzone. I hope yours continues to run great!

Great customer service/rma means nothing if the cards break easily.
 
This is akin to all the cheap items from amazon selling from china. Yeah, they have great return policy, but that just gives companies more leash to put out lousy products.
 
I've gone through 3 different hdmi 2.1 cables from Amazon that don't work with hdmi 2.1. Free and easy returns is why companies lie like this. Price it high and those that don't notice won't return it. I don't want this to derail the thread, but you get my gist of why I mention this.
 
And I don't mean EVGA is doing this because we don't know the cause, but free and easy rma provides a long leash for lackluster products.

I'm sorry to hear that. I've read on a different thread that somebody sent in their red lipped card that died and received a black lipped card back. Perhaps this newer version is less likely to have issues??

Mune had black lips too, it could be a bad batch? Lets hope we hear from EVGA soon

Good to know (about the black vs red lip being no difference). I guess I just have to hope that evga does something about this. What's the point of getting another card if it'll just brick again? Has anybody had a replacement brick?

AWK16

swunderly
Good to know (about the black vs red lip being no difference). I guess I just have to hope that evga does something about this. What's the point of getting another card if it'll just brick again? Has anybody had a replacement brick?

You mean right from Get-Go or after RMA

Sorry, I meant after RMA. Or are the RMA's a fixed solution?

I DNL Cyberpunk last night but afraid to run it if its a FTW3 KILLER !!!

If the purported 'global product management' staff remain resolute in not doing their job, by acknowledging these daily failures in any capacity(including denying an issue exists), the PCB changes in the next series of cards versus this one will serve as a makeshift announcement as to what the issue was.

swunderly

schulmaster
If the purported 'global product management' staff remain resolute in not doing their job, by acknowledging these daily failures in any capacity(including denying an issue exists), the PCB changes in the next series of cards versus this one will serve as a makeshift announcement as to what the issue was.

In the next series of cards? Do you mean the ti's? The 4000's? Or simply the next batch of 3080's?

It will not be any new production of existing SKUs, as the 'sua culpa' would be too severe. If and when the Ti's come, it would be risky*(edit below), but not surprising. After that, a virtual guarantee.

Some editorialized history: The EVGA 1080s cooled the GPU, but neglected secondary components that had historically never been a problem. Thermal pad application was demonstrated to be incredibly effective on their overheating/failing SKUs. Then came the 1080TI series, which I consider EVGA's tour de force in design. This was the birth of iCX, the multiple temperature sensors on both sides of the board on various components, and what reviewers could only call overkill thermal pad application.
EVGA, IMO, has coasted on that laurel and it has finally caught up with them. There was likely a negligible amount of preproduction testing done on their Ampere SKUs, made more likely by the fact that they had to delay their FTW launch in post-production due to realizing that the power filter configuration(CAP-gate) was insufficient.

As tacit as the response has been publicly, there has been and will continue to be internal communication aiming to avoid this from ever happening again, while avoiding the PR nuke that is anything recall-esque.
We as EVGA consumers may benefit from the bloodshed, as did those that waited until Pascal II.

*(EDIT): risky because the Ti would be positioned as an intermediary b/w the vanilla 3080 and 3090. A PCB redesign would disenfranchise 3080 owners, but would enrage 3090 owners, especially if the Red Light of Death ceases.

flg2010
I fear that any 3000 series card from evga is a ticking rma timebomb. I don’t see how they can fix this.
/*... */
Very discouraging

This is the failure that is most discouraging to me. Being the best company to RMA with does not offset the ample opportunities to do so.

Some of you guys are definitely leaving out details not to mention everyone now thinks it’s okay to just go ahead and go balls deep into overclocking a gpu without proper knowledge or cooling and continue to ramp these gpus to its limits until of course it breaks. Be honest here let’s count how many gpus that have been working fine compared to those that haven’t no one here is going to own up to their own mistakes.

The guy in that thread says he overclocked the gpu, to god knows what extent? He doesn’t mention how much he overclocked it see a pattern here? I want to know how many 3000 series died from not being overclocked at all, I doubt we will even know who’s telling the truth. Overclocking over the stock overclock if any gpu is not manufactured fault am I wrong here? When they sell you these gpus it’s whatever the clocks were specified yes or no?

flg2010
In my opinion, after reading this thread and others, I think the issue is related to: 1. Higher clocks/OC, 2. Power regulation, as in power supply or cables delivering unstable power and the cards not able to deal with that. 3. Poor cooler design

My card was failing between 55-68 degrees celcious at stock settings and in games like world of Warcraft (any settings), forza horizon 4 (at any loading screen).

Cyberpunk was running perfect I only black screened once there and only at a loading screen.

Also in EU, UPS is around 200€ a pop if you don’t want to send the card in 10-15 days and receive it God knows when. A complete disaster at this point. “The RMA is good” doesn’t mean anything, card should be working perfect out of the box with less than 50 hours gaming logged.