Aurora R8, bricked

I just benchmarked my 2080 with memory speed set to default and -502 which is lowest you can go and there was a difference of 3 FPS so basically no difference 

@DarbyDoo I do have Geforce experience installed, with the latest drivers. I also reinstalled the latest driver from Geforce, this seems to make no difference.

As for the video you suggest, I appreciate you try to help, but please do read the critical comments below it. A much more competent hardware tester has debunked most of the wild and unfounded claims:

https://www.youtube.com/watch?v=JIRfPlC15uc

Which is not to say I'm 100% sure your proposed solution won't work, I can't be sure of that. But there's strong reasons to question it, just saying. In any case, I will keep it in the back of my mind as a last resort. It would be a thing to go for if the machine were out of warranty. This machine is brand new, by principle I expect it to work correctly under perfectly normal load. 

Whilst at it, I'll share some support progress. Yesterday I raised a new ticket with Dell the Netherlands, and today they called me. Like the first time, they questioned the GPU problem, instead directed attention to irrelevant stuff like reinstalling Windows and running component scans. 

For some mysterious reason, they want to focus on the motherboard. They want to extensively test it and possibly do an on-site replacement. I'm not against it, but they cannot answer the question as to why they think this is related at all. Cynical me thinks they just want to avoid replacing such an expensive GPU once again. The support agent was actually pretty clear in saying that they cannot keep replacing the GPU. 

I agree, perhaps solve it by not placing a broken one in it? I insisted several times in explaining the failure behaviour of this problem: it happens after days or weeks, and seemingly at random. Once that first crash has arrived, it will keep crashing. It's a problem with a life cycle. You cannot reliably detect the problem with a standard component scan or hardware test but they still insist on this procedure.

I have no choice but to go through these cycles, so I'm now running some lengthy component tests. I'll let you know what happens next. 

---

@DromendeDaris wrote:

 

And...it's broken again. I had the first artifact crash on this replacement card, as seen on the screenshot. The artefacts first were constrained within the browser window, a few seconds later taking over the entire desktop and violently blinking.

How stupidly high is the failure rate on these cards to get two in a row that are broken? At this point, is it even a solution to get a 3rd try? 

 

---

My guess is one of 3 things:

- Replaced GPU was new, but from the same batch of bad ones

- Replacement GPU was a "refurbished" one, that is actually bad also.

- The problem is actually the motherboard.

@Tesla1856 

I have no basis to know which of those options is most likely, but it doesn't hurt to speculate, I suppose.

The odds of having two failed cards in a row assuming RMA rates of 1% - 3% (I have no idea if these numbers are accurate but it's what I read) in the most "likely" case would be roughly 1 in 1,000, up to 1 in several thousand. 

Possible, but unlikely. Therefore, I agree that the likeliness of having two failed ones in a row can be spectacularly higher if indeed that 2nd card is from a failed batch stocked up at Dell/Alienware where the fail rate is into the dozens of %, or maybe just the entire batch is broken. As we know, there's a delay factor in the failing. If you'd stock up a 100, it will take months to know what the real world fail rate is.

I don't know if Alienware stocks up or does just in time ordering from Nvidia. What I can share is that in tracking the repair status of my first card, it was in status "ordering components" for about two days. That seems pretty short to me if it would be ordered from Nvidia directly. I therefore assume it's an internal order. I could be totally wrong, of course.

I don't understand the relation with the motherboard. None of the information I have seen suggests it as related, yet both Dell today on the phone and now you consider it a possible factor. Please educate me why you think it may be a root cause?

---

@DromendeDaris wrote:

 

I don't understand the relation with the motherboard. None of the information I have seen suggests it as related, yet both Dell today on the phone and now you consider it a possible factor. Please educate me why you think it may be a root cause?

---

These "bad" cards should be tested in a different (known working) machine. That would answer some questions, and yes, people do it all the time. It's basic trouble-shooting.

If it's not the software, it's hardware.
If 2 new cards fail to work, you start looking at other things.
The motherboard is next. Motherboards (like video-cards) fail all the time.

Although the 2nd card failing is a huge dissapointment as was the initial support contact, Dell won back some credit today. Whilst I was running ePSA tests that ultimately fully passed, Dell the Netherlands contacted international management for guidance on what to do in this hairy situation.

Their answer was simple: replace the GPU. So it's going to get replaced again and this whole motherboard investigation is cancelled. They will do the replacement on-site, so that unlike last time, I will not lose the machine for weeks.

So finally, the problem is taking serious, and dealt with in a customer-friendly way. 

Which up next of course leaves me wondering about the odds of a 3rd card being good or bad. That thinking may be too negative, but you can imagine I'm thinking it after 2 failed cards. I offered being open to discussing a non Founders Edition 2080TI to reduce the risk for both parties. I don't expect that to happen, it's just an idea.

Will let you know what happens next. 

Update: 5 days ago, I received on-site support where the 3rd card was installed. 

When the engineer opened the box containing the GPU, it had a huge label on it saying "REFURBISHED". I'm obviously not happy getting a used card, and worry why it was sent back by another customer in the first place. I was promised a new card of a "good batch", not this.

I thought about how much to protest against it, but decided to give in for now, as my main priority is to simply get a working card and to stop wasting so much time on this computer. 

Refurbished or not, so far the card has worked well, not a single glitch in 5 days of usage. I hope it lasts, remember how my first card had 6 weeks of problem-free usage before it crashed, so I will only be able to tell in the longer term.

nice PC with Gen 8 or 9 CORE processor. Z370 is  PRIME  (gen 9 dumps 95watt TPD)

the PC has 2 GPU chips, Intel UHD Graphics 620 or 630 (gen 9 CPU)

so is the PC overheating,seem to me the GPU card is, so remove it and run the PC on the 6n0 chip port

remove NVIDIA GeForce RTX 2080 (or TI)

and for sure always run cpu/gpu temperature monitors

is the PC in a hot room or worse a hot closet or CABINET? we can't no environments, no photo's

It is overheating or the RTX is bad.  that is what I'd assume.

if the PC runs ok on the 6n0 GPU engine, then buy Nvidia GTX1050 and watch it run perfect, and very cool.

make sure  all fans are working. PSU, and GPU  card. and more if more there IDK'

this PC should have had the 850 W PSU option not the 460w.

that is my help.   based on no testing, yet.  (get the GPU card out of  the PCI-e slot first

and learn to use thermal tools.  Hwinfo64.exe portable and free and clean is best. try it.

I appreciate all of these posts... I just created a brand new post in this support group re: my 3 month old Aurora R7 (Nvidia GTX 1080 card) and how my new PC is literally bricked. Was playing The Division 2 (after playing it for weeks with zero issues) and my PC literally froze up, played a garbled loud scrambled noise and I had to manually shut down the PC... for it to never get to boot into Windows 10 again. A full day of trying, using BIOS to boot from my USB recovery stick (which did NOT work) and now my PC is at Micro-Center since they installed a new SSD into the NVME/M.2 spot a month ago -- have no idea if my SSD failed, if my graphics card failed, or what. All I know as I am writing this is Micro-Center said they've had to order a new motherboard from Dell.

This has been a TERRIBLE experience and I am a first time PC owner. Not sure I did the right thing. 

Thanks for all this information, I was just about to buy one with the i7 9900k and 2080ti.

I Bookmaked this page so let us know what going please I look foward to see it.

anymore issues?

Important update. I manually reinstalled nvideas latest driver using "clean install". Also, I did not install GeForce Experience. So far no more space invaders and crashes - before 100% reproducible in Futuremark Time Spy.

Crossing fingers...