NVIDIA on the Cause of RTX 2080 Series Card Failures

Me>?... Juang was right. If you bought two you may get lucky maybe you will only have to RMA just 1. You see, the more you buy the more you save.
 
Last edited:
Um... Yeah, just like the one that burst into flame while simply surfing web pages...

Kyle, longtime reader here. Love this site, best source of unbiased reviews on the planet.

Original: We found that people that are having the most problems are putting these cards under too much stress. People using the cards for general productivity apps, web browsing, etc., will likely not see any problems.

Adendum: Some web browsing might cause similar issues. (Thanks go out to [H]ardOCP)
 
What about cards with custom PCBs like the MSI Gaming X and Asus ROG Strix? Should they be ok?
 
I feel bad for those stuck with such pricey ticking time bombs. To say the least my enthusiasm and confidence would be really low about now. In fact if I owned one of these I might be tempted to return it for a refund until these issues get identified and solved. I used to be all over bleeding edge hardware on release day. For the most part new gear is uneventful and introductions smooth. I can remember back almost 2 decades ago when PII and PIII CPUs ruled. Anyone remember all the motherboards that got sold with really cheap capacitors on them. We were seeing tons of bulging leaking parts and of course erratic failing boards. I just can't help but wonder if these same accountants are buying components again? Who in their right mind would risk a recall over a few cents of cheaper parts? Scammers and fake parts are rampant though. I restore audio equipment and only buy parts from reliable vendors. I'm sure its easy enough between the final build revision and doing production runs that bad stuff could sneak in. Shouldn't random samples and high load testing be done during production though?
 
Low yields and no quality control on the 2080ti's. Is that the gist of it or am i missing something.
 
Didn't they just recently say "0.01% RMA Rate"?!

That is every car company in the world, it's a manufacture defect and they are like nothing to see here.
Then later on they come clean like: yes there are problems we will fix it for you.

To connect that together: Nvidia started working with car companies and took some hints from them on how to handle things of this nature.
 
Well, just goes too show, Nvidia screws up just as well as anyone else. That said, I hope they are overnighting anyone that has this issue and can prove it a new card. I mean for $1200, nothing less should be expected. (I know how I would feel if it were me but then again, I would feel the same way if it were just a $300 card.)
 
There's no escaping statistics at high volumes; I have had bad parts make it past a triple-sort. :mad:

Of course it can happen. But I'm thinking (1) the production volume on all these boards will be limited by the small market for $800+ cards, and (2) a part has to be crucial and bad in a particular way to cause such a catastrophic event. So the probabilities get smaller and smaller, to the point where 'bad parts missed by testing' as an overall answer seems strange to me. Unless combined with unstated factors such as "bought most parts from new low-priced source".
 
Last edited:
They should replace no doubt and any other damages done but doesn't warrant a refund. While rare high powered electronics do catch fire. I had a MSI 780 catch fire years ago. I opened a rma ticket and got a new card in a week over Thanksgiving.

Exactly right: it is rare. And potentially catastrophic. And very damaging to a firm's reputation especially with a newly-released, highly-hyped, ultra-dollar product. So EVGA would be wise to do everything possible to mitigate the effects of this rare event. I never said they have to issue a refund; but under these circumstances it might be smart to do so. That would promote the idea that it's a one in a million chance. Unless they expect a lot more to fail like this...
 
Anyone remember all the motherboards that got sold with really cheap capacitors on them. We were seeing tons of bulging leaking parts and of course erratic failing boards.

*nods* lol
Here's another great example. Once upon a time, a capacitor company thought it had bought on the black market, the secret recipe to the electrolytic solution in a high-end competing company's capacitors... This company went on to produce god knows how many capacitors using this recipe, thinking they were going to be rolling in the cash as well. Not too long down the line, products that features these caps started to fail prematurely. They ran fine up until then, but eventually the electrolyte boiled off and (presumably) lead to thermal runaway, causing them to burst or leak.

I didn't outright say it, but that was what I was referring to. I had a Shuttle SV25 system (FIC FV25 mobo) and those 5 Green & Gold caps all buldged and went bad. The sad thing is, I barely used the board. It sat for most of its life, getting only a few hours of total use time (I completely bodged the cap replacement due to soldering inexperience, though I still have the board for whatever reason lol).

That's why I don't find their excuse all that unbelievable, and why back on page 2 I had argued that it doesn't necessarily need to result in a DOA card, that it'll take time for these "escaped" parts to degrade enough for there to be issues.


Anyways, my money is on nV claiming it was an AMD loyalist working against nV by sabotaging their cards with these "escaped" parts... :pompous: lol
 
:::cough:::

2018-11-15_17-14-40.jpg
 
*nods* lol


I didn't outright say it, but that was what I was referring to. I had a Shuttle SV25 system (FIC FV25 mobo) and those 5 Green & Gold caps all buldged and went bad. The sad thing is, I barely used the board. It sat for most of its life, getting only a few hours of total use time (I completely bodged the cap replacement due to soldering inexperience, though I still have the board for whatever reason lol).

That's why I don't find their excuse all that unbelievable, and why back on page 2 I had argued that it doesn't necessarily need to result in a DOA card, that it'll take time for these "escaped" parts to degrade enough for there to be issues.


Anyways, my money is on nV claiming it was an AMD loyalist working against nV by sabotaging their cards with these "escaped" parts... :pompous: lol

The funny part of the whole Capacitor fiasco is that someone stole the tech, moved it to a different company, and didn't know that they didn't know the whole formula until they started failing.

There was an additive to make the electrolyte less corrosive, and it got left out, so the electrolyte ate the seals, cans, and leads, making them leak and fail.

I repaired a dozen or so Dells for just that issue. :)

All the caps I've had fail at home were just really old, or power surge damaged.

https://en.wikipedia.org/wiki/Capacitor_plague


EDIT: I was sitting here browsing several sites, and I heard a noise as I started playing a video; it was the fans coming on on my RX480. :)

I've been on it all day, and the fan didn't start up until I started a video from YT.

That's pretty cool.
 
Last edited:
I work in manufacturing. I'm going to leave it at the tinfoil hattery isn't unfounded.

I wish I could share your idealism. And quite the contrary, cheap parts + don't pass the savings onto the consumer = more profit.

Can't see how you figured idealism! It's economics: putting risky low-value parts into high-dollar products, where profits are large and brand is crucial to selling, makes no sense economically. Neither EVGA nor NVidia got where they are today by being foolish. That's why I find it surprising that "parts quality", where a batch slipped through the statistical sampling, is the overall issue.

To put it another way: I'd suppose they'd be buying parts so good that a final round of testing - after what the vendor did - would rarely find a defect. To have a failure in the final check process coincide with some lots of bad parts is... surprising. Especially, again, in $$$ halo boards where a round of serious failures will cost them hugely.
 
This isn't going over very well have about 40 hours of gaming under my Evga 2080 played Kingdoms come deliverance Monster Hunter COD black ops 4 Hunt Showdown and Pathfinder Kingmaker for 5 hours today its going pretty good so far had 2 lockups with Hunt Showdown an COD lockups went away with updated patches. I uninstalled about four programs a Korean Crytek steam rep told me to free up ram with Hunt showdown.
 
I know you know what passive cooling is so I'm not going to insult your intelligence. Your card is being air cooled. I was responding to a card being passively cooled.

Right. Sorry I didn't pick up on what you meant in that context. When discussing video card overclocking anything short of heat sinks touching the VRM is often considered passive.
 
This isn't going over very well have about 40 hours of gaming under my Evga 2080 played Kingdoms come deliverance Monster Hunter COD black ops 4 Hunt Showdown and Pathfinder Kingmaker for 5 hours today its going pretty good so far had 2 lockups with Hunt Showdown an COD lockups went away with updated patches. I uninstalled about four programs a Korean Crytek steam rep told me to free up ram with Hunt showdown.
Is this why his forum title is "Ignore Me"? lmao But seriously, that was rough to read :(

On a non-jackass note... err, well kinda... Is it a "good day" in nVidia Land when you only have 2 lockups while gaming? :confused:
How bad WERE things for you, that 2 is qualifiably good for you? :(
 
Can't see how you figured idealism! It's economics: putting risky low-value parts into high-dollar products, where profits are large and brand is crucial to selling, makes no sense economically. Neither EVGA nor NVidia got where they are today by being foolish. That's why I find it surprising that "parts quality", where a batch slipped through the statistical sampling, is the overall issue.

To put it another way: I'd suppose they'd be buying parts so good that a final round of testing - after what the vendor did - would rarely find a defect. To have a failure in the final check process coincide with some lots of bad parts is... surprising. Especially, again, in $$$ halo boards where a round of serious failures will cost them hugely.

You're right, it doesn't make sense. You aren't going to get a disagreement out of me on that.

However, I've also been in this business long enough to know that sometimes, the people at the top get a sudden urge to chase pennies, even though there are perfectly good piles of dollars neatly bundled up and awaiting their new wallet homes. Why? So at the end of the quarter, they can proudly say "look at how much money I saved the company!" and/or "look at how i managed to boost profits!"

Logic would dictate that you would put nothing less than the best in your flagship product while still retaining profitability, and keeping quality issues to as low a minimum as possible. After all, it's the thing that is defining your brand, your image, and establishes how the world views your products. Not every single unit is going to be 100% awesomesauce; it's inevitable that some are going to have issues. However, great support and customer service are what will continue to polish up your brand, and make it continue to be seen as a great product, and a trusted brand.

How pissed would you be to find out not every company views it that way? And some of them are the biggest names in their respective industries?
 
Of course it can happen. But I'm thinking (1) the production volume on all these boards will be limited by the small market for $800+ cards, and (2) a part has to be crucial and bad in a particular way to cause such a catastrophic event. So the probabilities get smaller and smaller, to the point where 'bad parts missed by testing' as an overall answer seems strange to me. Unless combined with unstated factors such as "bought most parts from new low-priced source".

The low volume part is the GPU; the caps and resistors will be high-volume and normal probabilities apply. When you ship 10k pieces a day a bad part can get deep in the chain before somebody has even read the report. Of course, this is known and can be accounted for, but clearly wasn't.
 
The low volume part is the GPU; the caps and resistors will be high-volume and normal probabilities apply. When you ship 10k pieces a day a bad part can get deep in the chain before somebody has even read the report. Of course, this is known and can be accounted for, but clearly wasn't.

*Cough* Bumpgate *Cough* Or NForce 3 working only with Nvidia cards. Yeah, they do things on purpose, I highly doubt this was simply a case of missed issues. *GPP* Long story short, this is just wrong and Nvidia needs to fix it straight up through all vendors, not just the ones they sold directly.
 
I think this presents an excellent opportunity to get NVIDIA stock at a nice discount. I bought 3k worth this morning.
 
Yeah, AMD dropped 6% after close too. Must be something going on in the financial world that investors aren't liking.

Nope. Revenues down. cards/mining and self driving cars not coming to a showroom quite as soon as people thought.
 
Yeah, AMD dropped 6% after close too. Must be something going on in the financial world that investors aren't liking.
I have been advised by someone extremely in the know to keep all my assets as cash at this time. Tread lightly. I asked about blue chip stocks even. He said no.
 
FYI: eTeknix just reported that NVidia has dropped the 2080 TI from their on-line store.

We'll see how this shakes out, but it indicates that the problem is not with inadequate testing of jellybean components - which as I've been saying seemed unlikely.
 
Last edited:
FYI: eTeknix just reported that NVidia has dropped the 2080 TI from their on-line store.

We'll see how this shakes out, but it indicates that the problem is not with inadequate testing of jellybean components - which as I've been saying seemed unlikely.
Haven't we been talking about this for a couple of days now in various threads?
 
My 2080 Ti back-plate gets hot as heck (and cam temp alerts all the time) since moment I installed it have been hot - the fans sit 41% with little bump 43% once in a while - Until last night when I found EVGA Precision X1 beta, unlike other tools used to use (2080/2080 Ti fan control grayed out) you can control the fans (have them at 100% now).

Now my 2080 Ti (launch day Founders) sits 62 or 63c (even in FurMark) with clocks staying above 1905 and memory at 1750. The lock ups in couple games are gone, frame rates are up little bit as well.

I realize that 82+c is within spec, however this 2080 Ti seems to much prefer the lower temps, and now it just works... just hoping it doesn't flame up, lol...

EDIT:
Has anyone has their 2080 or 2080 Ti fail with it on custom loop?
 
Unless Nvidia is pushing the limits on this memory, isn't this ultimately Micron's fault?


.
 
It's already been shown that the majority of the space invader issues don't seem to be a temp issue. It could be the memory, but no one has shown exactly what it is. At this point, it seems as though nvidia has corrected the issue though, as we aren't seeing many reports anymore.

Myself, i'm still running my 2nd shipment batch 2080ti FE from Nvidia with no issues, and at this point the thing has been hammered.
 
I waited til last month, bought evga 2080ti ftw3 ultra something or another at microcenter in Dallas. I went home, DDU'd, removed 980ti installed 2080ti, booted, installed driver, and have been hammering it since. It has exhibited no faults.
 
Back
Top