RTX 2080 Ti RMA Rate "Under 0.01%," GTX 1080 Ti Stock Almost Gone

I've been following this story for about 2-3 weeks now. About a half dozen sites have posted similar threads asking owners to tell their stories. Majority of posts are speculations, comments, questions, and unrelated posts. I've also noticed how some of the reports are being repeated in multiple threads. It's been a challenge to discern the facts from the noise. Ironically this post lends itself in that direction. I don't doubt there's been failures(they happen for all tech) but the lengths of the threads are very misleading since only a few of the posts actually pertain to the question and as I stated, those stories are repeated across multiple sites. After it's all said and done that .01 is looking more accurate than not, that is unless more reports do come out.

edit: I did want to add that I totally believe about the 1080TI stock. Once I cancelled my 2080TI order I began considering ordering a 2nd Strix 1080TI. After all, I paid $759 for my 1st. When in stock, they've lately been averaging $1300-1500! Now that's ironic pricing!
 
his product launch is strange and confusing.

First of all, why is any Ti released along with the regular product right away? In this case, to drive up the price.

The 1080 was released like a year before the 1080Ti. It made sense.
Now it's like the 2080Ti is what the 2080 used to be in spirit, but because NV seems to need the money or can drive the price for lack of competition, they release what SHOULD be the 2080 as the 2080Ti and expect us to open our wallets in a major way.

I'm of the "screw it" variety, keeping my 1070 going on a new G-SYNC screen to deal with framerate dips and waiting another 2 years.

2080ti is a Titan in spirit, not a mid range die. If you compare the ti to the original Titan X Pascal position it’s nearly identical and the price is the same, except we don’t get a god awful blower.

As per the 0.01% rma rate. I have a hard time believing the public at large would only mess up (and rma) only 0.01% even if the cards were 100% working perfectly. Human error has to be at least 1%.
 
Wonder how many of the 13% actually own the card.

To counter your conspiracy theory, Cybereality has gone through 3 ti cards. That pendulum you are swinging goes both ways.
 
it's Nv..would not be first time they "squash" information so as they appear everything is "normal" (nothing to see here move along) for example many of the G80's, G90s, some of the 200s-400s-600s or even that 970 only having X memory available for use.

They will do "whatever they can" to spin information however possible even when they know/knew what the problems were.

To me this seems very "normal" when it comes to Nvidia, problem discovered, they try to make you "look away" instead of admitting "in full" what the issue was and what they are doing to fix it type deal.
 
I mean..the xbox never had a red ring problem...you were just using it wrong.
Funny story, I was at Microsoft's office with a bunch of game reviewers before the original XBox launched. I was the only hardware guy there. An engineer came in and presented us the internals of the XBox and discussed cooling, and I questioned him pretty hard on it. I told him they were going to have long term issues due to cooling not being what it needed to be on that machine. He got kind of huffy with me, and basically walked out of the interview. And then I was never invited back to Microsoft again. LOL.
 
Either nVidia has shut down the studio so they can figure out what's going on in terms of QC, or else people are ready, willing, and able to pay $1000+ for a GPU. I lean toward the latter being true and that means it's a sad day for gaming, IMO.

Nvidia proved that was the case when they launched the Titan years ago. The only way to bring prices down is to not buy it, but people have been. I don’t blame Nvidia for this. If people are willing to pay $1000 USD for a video card, then that’s what people are willing to pay.
 
It is FE cards that are having the higher returns, most AIB cards have better coolers on them. Nvidia even admitted they are seeing issues cropping up, which means that it is above normal, otherwise they wouldn't be saying anything.
It seems any card that uses the FE reference PCB is more likely to have an issue. To be clear many AIB are using the reference PCB and just putting an aftermarket cooler on it.
 
2080ti is a Titan in spirit, not a mid range die. If you compare the ti to the original Titan X Pascal position it’s nearly identical and the price is the same, except we don’t get a god awful blower.

As per the 0.01% rma rate. I have a hard time believing the public at large would only mess up (and rma) only 0.01% even if the cards were 100% working perfectly. Human error has to be at least 1%.

No not really. It's 11GB memory. The Titan used all available memory channels.

This is an interim release only. (Until 7nm gets knocked out) There's no time to release this, wait a year, then release 7nm. The 2080 series will be quickly replaced in 1 year.
 
No not really. It's 11GB memory. The Titan used all available memory channels.

This is an interim release only. (Until 7nm gets knocked out) There's no time to release this, wait a year, then release 7nm. The 2080 series will be quickly replaced in 1 year.

They were both very close to full die. Using all memory channels for 1GB 9% more ram isn’t really a differentiator... if they called the 2080ti a Titan no one would blink. We can bicker over if it’s just slightly off from a Titan but the guy I quoted made the 2080ti sound like a mid range die which it definitely is not.

There’s always something better just around the corner.

The quarterly report for this quarter should be interesting. Maybe it’ll give us some insight to what is actually happening.
 
Last edited:
It seems any card that uses the FE reference PCB is more likely to have an issue. To be clear many AIB are using the reference PCB and just putting an aftermarket cooler on it.

Might be but the aftermarket coolers are keeping the gpu temps down a lot more than the fe stock cooler. Some cards are loading in the mid 60's compared to 80c+ for the nvidia cooler. If heat is indeed the issue and other coolers are doing a better job on the same pcb then its down to nvidia cooler more so than the pcb.
 
Might be but the aftermarket coolers are keeping the gpu temps down a lot more than the fe stock cooler. Some cards are loading in the mid 60's compared to 80c+ for the nvidia cooler. If heat is indeed the issue and other coolers are doing a better job on the same pcb then its down to nvidia cooler more so than the pcb.
I don't think that you're looking at it correctly. From what I read from a couple sources it's the actual layout of the PCB that's contributing to some of the heat issues as their components in places that can't be adequately cool or components on the bottom of the PCB underneath other components or something like that. So yes the aftermarket coolers may do a better job but the underlying problem appears to be at least partially to blame on the PCB itself.
 
Hate to be this guy, but may I ask where 13% came from with any sort of statistical credence? How large is the sample size we are looking at (I know there are many members on [H])? And is it just the vocal people upset in forums created for enthusiasts or the majority? Do we have any sort of control for that? 13% seems insanely high to me and much more of an outlier than statistical relevence...

Well said. 13% doesn't make any sense.

So much hate on the price of this card with a lot of kids not being able to afford it I wouldn't be surprised if some if not most of the reports are false.
 
LOL.

0.01% is an industry standard level, For Extremely Reliable Hardware.

I have PCB's in the field that HAD to maintain that level, or Action Was Required. (Medical equipment, had to last 10 years +)

There are 300k+ of them in the field, and that spec is quoted over the entire lifetime.

There is NO video card that will ever hit this level; it's too hard to do with stuff that doesn't run 85°C.

There are going to be more Fans fail than that level, lol.

This report is fucking laughable to anyone in the electronics industry.

I don't care Who they are, 0.01% is ridiculous; ASK any EE you know.
 
I love how the people discussing this in this thread are commonly calling people who disagree with them delusional. It's as if there must be some overriding correctness and only one person knows it but everyone thinks the are that one person.

Guys yea I'm sure the failure rate OVER THE LIFE CYCLE is greater than .01%. That number is pretty damn low. BUT, I also suspect the current failure rate is higher than that by a fair margine and they are trying to say of this chipset across all manufacturers the failure rate is less than .01%. And only time will tell if that is true.
 
That is why I used the words seems like, I only use the polls here and really don't read too much news as it's just one big echo chamber (5000 articles for one event).

The reality is this launch has been a shit show. Nothing in stock for ti's over a month post launch, after being delayed a week, and still no sign of RTX features on SotTR. Maybe that will change come BF5's launch but my instincts say it won't.

2080's languishing on shelves as well, very little product movement at my local stores anyway.

Circling back, we have reports of failures, even in these very forums. What a mess. Any time someone says you need to redefine performance, they're shoveling it to you.

I don't even really take the H poll for the 2080 ti failure as a scientific. Generally speaking, if you don't have a problem, you probably won't be searching for that topic. If your new car is working flawless you probably aren't going to search for "camaro 2019 steering wheel vibrates."

That being said, if there is a problem then it needs to be addressed. But after being around the internet so long every problem tech is a "-gate".
 
I would suggest just the opposite. I've seen an absolute deluge of negativity surrounding the new nVidia line-up mostly due to cost. You really think these kids are innocent of making false reports to an "open to the public" survey? Highly likely. In fact, anyone who doesn't think this would happen is delusional. I am sure many of the reports here and elsewhere are false.
I'm pressing F for sad victim Nvidia being bullied by poor forum goers.
 
0.01% is 100% accurate, but that's because 99% of them sit on the shelves at that horrific MSRP. Can't break em if you don't buy em.
 
I would suggest just the opposite. I've seen an absolute deluge of negativity surrounding the new nVidia line-up mostly due to cost. You really think these kids are innocent of making false reports to an "open to the public" survey? Highly likely. In fact, anyone who doesn't think this would happen is delusional. I am sure many of the reports here and elsewhere are false.
Do you have any facts or are you being somewhat delusional? :)
 
I would suggest just the opposite. I've seen an absolute deluge of negativity surrounding the new nVidia line-up mostly due to cost. You really think these kids are innocent of making false reports to an "open to the public" survey? Highly likely. In fact, anyone who doesn't think this would happen is delusional. I am sure many of the reports here and elsewhere are false.

There is an absolute deluge of negativity around every graphics card release from both AMD and nVidia, remember the founders tax first introduced with Pascal? AMD's heat and power issues? Price issues always?

Your reasoning is rejected, it is as anecdotal as the reports of failure and coupled with the constant stream of negativity from both sides against any new product any false reports would also exist in likely equal numbers with every GPU launch.
 
There is an absolute deluge of negativity around every graphics card release from both AMD and nVidia, remember the founders tax first introduced with Pascal? AMD's heat and power issues? Price issues always?

Your reasoning is rejected, it is as anecdotal as the reports of failure and coupled with the constant stream of negativity from both sides against any new product any false reports would also exist in likely equal numbers with every GPU launch.

I remember fermi.

nvidia.jpg
 
The very poll here on hardocp dictates that it is no where near 2% or 4%. It is much higher.
A poll on HardOCP is by no means scientific and there isn't even close to enough data (yet) to reach any type of firm conclusion, IMO. That being said I do find stuff like that interesting. The board partners and nVidia have the data and know the truth. Whether or not they convey that is the true question. They do have motivation to lie. And corporations are inherently corrupt. Slaves to the almighty dollar. That's all they care about. So with that in consideration, they are probably lying. However the AMD "people" are lying also. It's all a big game. You can't believe what you read and you can't believe a lot of what these companies will tell you.
 
The percentage could be very low depending on how much are sold ;)
A poll on HardOCP is by no means scientific and there isn't even close to enough data (yet) to reach any type of firm conclusion, IMO. That being said I do find stuff like that interesting. The board partners and nVidia have the data and know the truth. Whether or not they convey that is the true question. They do have motivation to lie. And corporations are inherently corrupt. Slaves to the almighty dollar. That's all they care about. So with that in consideration, they are probably lying. However the AMD "people" are lying also. It's all a big game. You can't believe what you read and you can't believe a lot of what these companies will tell you.

If you take a low percentage that must mean that all of the people with bad cards are on forums :)
 
I don't think that you're looking at it correctly. From what I read from a couple sources it's the actual layout of the PCB that's contributing to some of the heat issues as their components in places that can't be adequately cool or components on the bottom of the PCB underneath other components or something like that. So yes the aftermarket coolers may do a better job but the underlying problem appears to be at least partially to blame on the PCB itself.

Partly maybe but the FE card has virtually zero airflow to the pcb, it has a vapor chamber with a dinky little aluminium heatsink attached with a sealed off shroud meaning the hot air doesn't get pushed away from the card as quick as most of the aib cards. It doesn't have a lot of venting room compared to aftermarket cards either.
 
Now some more info:


For those who don't want to/can't watch:

2080 ~ 0.17%

2080Ti ~ 1.4%

He said that the sales numbers are approaching 4 figures for 2080 and a bit less for 2080Ti, so the sample size is reasonable.

He also has RMA rates on 1000 series:

1080 - 7.1%

1080Ti - 4.6%

Keep in mind ofc that those are on market for much longer and there's a lot more time for shit to go wrong.
I honestly don't know who to trust or what to believe at this point. It all seems like a bunch of hot air from both sides.
 
Now some more info:



I honestly don't know who to trust or what to believe at this point. It all seems like a bunch of hot air from both sides.

Thanks for the post.

Considering what a legend this guy is I do trust his insights on the matter. Also considering all the hype in the various forums I am a bit more speculative.

I tend to spend a lot of time reading reviews from around the world on GPU releases. If factoring in the number of reviewers vs. consumers vs. cards sold it would still be unusual, but not impossible, for a reviewer to get a bad card. That being said, so far I haven't found any. Estimating to compare 2-3 dozen reviewers vs. 1000's of consumers could easily be misleading. It would just add a bit more credibility to this if we encountered a professional with direct involvement with the product straight from opening the box.
 
his product launch is strange and confusing.

First of all, why is any Ti released along with the regular product right away? In this case, to drive up the price.

The 1080 was released like a year before the 1080Ti. It made sense.
Now it's like the 2080Ti is what the 2080 used to be in spirit, but because NV seems to need the money or can drive the price for lack of competition, they release what SHOULD be the 2080 as the 2080Ti and expect us to open our wallets in a major way.

I'm of the "screw it" variety, keeping my 1070 going on a new G-SYNC screen to deal with framerate dips and waiting another 2 years.

Same boat. I play older games at 1440p. Currently working on Metro Last Light Redux after finishing 2033. SLI with 1070 @ stock speeds is giving me 70-115fps with 2xSSAA on. Just need a g-sync monitor now.
 
Hate to be this guy, but may I ask where 13% came from with any sort of statistical credence? How large is the sample size we are looking at (I know there are many members on [H])? And is it just the vocal people upset in forums created for enthusiasts or the majority? Do we have any sort of control for that? 13% seems insanely high to me and much more of an outlier than statistical relevence...
Kyle said exactly what it means: 13% of the people visiting that discussion thread have reported a failure. If you're looking more into that than what it represents, then that's your problem.
Haha. Dam, so might be easier to ask which companies haven't you pissed off? :)
Like in Futurama when Leo Wong mentions they own so much stuff on Mars that it's easier to brand everything they don't own.

upload_2018-11-6_9-44-1.png
 
Now some more info:



I honestly don't know who to trust or what to believe at this point. It all seems like a bunch of hot air from both sides.

well, two things, First:- De8auer works for Caseking, He isn't going to say anything that will affect their sales. Second thing, Caseking don't sell FE cards and the problems seem to be mainly with the FE versions. So his numbers tell us nothing really. Besides,I never fully trust numbers from anybody trying to sell me something.
 
Back
Top