spider1701

kevinc313
 

I agree it is clearly a highly abnormal load situation that any normal stress testing regimen would not replicate, exposing a weakness in the card, possibly vram temps and/or vram power delivery.  
 
BTW, the ancient ATI tool 3d render will let you run 6000fps at about half power.  People use it to test max stable core clock.

cool... at least is is on 50% power :-) (I am tempted to try it.. any download link for it?)
 
What makes me wonder Amazon claims it is not their fault yet they are quickly deploying patch limit FPS in menu :-)
this will be interesting to follow up in the future what outcome there will be for them.

Not Amazon's fault for sure. But also for sure Amazon don't want their game to be the one exposing graphics cards design flaws. As you can see a bunch of people has said to not playing NW because they naively think that's a way to protect their cards. So for such an easy patch Amazon will sell tons more of game copies. By the way for those thinking that not playing NW is a way to protect your cards, it doesn't matter, because there will be another game, another software that will reproduce the issue in the future unless AIB addresses this issue.

Just caught up.  This whole ordeal is nuts, but so is everything else in this world right now, right?
 
I think there are a few takeaways at this point:
1) As the failure is occurring on other brands/cards as well (AMD cards included), It's not just EVGA.  In general, It just happens EVGA sells the most cards out of all AIB's out there naturally you will obviously see more EVGA failures (This does not give EVGA a free pass).
2) nVidia should have given stricter instruction to AIB MFGR's (as someone said previously)
3) FOR THE MOST PART: This is the first time in history we've have a Single GPU pulling well over 300+ Watts in a game off a power supply (You're talking 450-500W).  For reference, All previous high end generations maxed around 250W or so **generally speaking**.  If you have SLI or CrossFire of an older generation, there is no comparison.  You're talking about multiple cards at this point compared to a single card, and the way the power is driven is different.
 
4) This brings me to my last point:  We should find out what Brand/Model PSU's people are using.  These cards dying may not be related to the PSU at all, but it would be interesting to add the PSU's to the list of everyone with a dead GPU.  We all know RTX 3000 series exposed weaknesses in certain PSU's when 3000 series launched.  So why not?  I understand the Power Delivery is handled by the VRM and parts on the Card, but the Delivery comes from somewhere.
 

spider1701

kevinc313
 

I agree it is clearly a highly abnormal load situation that any normal stress testing regimen would not replicate, exposing a weakness in the card, possibly vram temps and/or vram power delivery.  
 
BTW, the ancient ATI tool 3d render will let you run 6000fps at about half power.  People use it to test max stable core clock.

cool... at least is is on 50% power :-) (I am tempted to try it.. any download link for it?)
 
What makes me wonder Amazon claims it is not their fault yet they are quickly deploying patch limit FPS in menu :-)
this will be interesting to follow up in the future what outcome there will be for them.

Here:
 
https://www.techpowerup.com/atitool/
 
Just run the fuzzy cube and keep cranking up core clock until you get a crash/instability, can usually go well over 2,200mhz.   Might want to have GPU-Z or HWiNFO to keep an eye on stuff.  Proceed at you own risk but I've never heard of a problem with it.  It is a good indicator of what your chip is capable of.

spider1701
 
I really hope it will be clear if it is problem of HW design, Firmware or Drivers.
Best case scenario would be firmware as it could be fixed. If the HW design will be proven to be purposely flawed (cheep component etc.) than it will be lot of pain for EVGA.
On the other hand their HW (PCB) were certified by NVIDIA so hard to tell who to blame if someone :-)
 
The rest of the affected cards on amazon forums seams to survive the experience (hard crashes etc. but they survived).

Yes, there could be a FW or SW fix of course, I do not doubt it will, but that only will be fixing (in part) a major HW flaw. You could expose the same in other scenario. On the other side, it is true that this cards are extremely powerful, and with 3 - 8 pin connectors, they (AIB brands) may have risked too much. Maybe the only card that should have had 3 connectors (and support the amount of w we're talking about) was the kingpin.

Zanza741
Supposedly the problem is the fan control chip. That is extra from EVGA on card. (iCX3) Normally you do not need such a thing to regulate the cards via GPU.

What about warranty, is such a defect covered?

From what I heard, EVGA is covering this issue. (Their RMA reason section was updated to have "New World Game Issue" as an option.)
 
Can someone point me to where there's a collection of PCB revisions? I saw that people were saying that only certain revisions had this issue and wanted to know where they got that information.
 
I also monitored my 3090 FTW3 HC and it got to a max of 47c while playing for 3+ hours. I'll check the memory and (TJunction?) temps later today but does anyone know if the HC lacks that chip? Wouldn't think that a 3090 HC would have/need a fan regulation chip.

[Deleted]

For what it's worth, RTX 3080 owners are also reporting this. I had something very similar happen to me while playing FFXIV on my 3080 FTW3 Ultra so I'm assuming it's a similar problem. Theoretically, setting a maximum FPS globally should stop whatever application/game from drawing obscene amounts of power during unlocked loading screens, right? I'm about to RMA my 3080 to EVGA today but in the meantime, I'll be setting a global FPS cap to my monitor's maximum refresh rate. 

tamerlee
As a 3090 FTW3 owner that has played New World heavily over the last few days, this is terrifying.

Can you tell us what revision your 3090 is? (Or when you were able to buy it?)

jaketeck
Interesting. These boards have shunt resistors and power monitoring all over the place. I suspect that there may be a path that doesn't have power monitoring and that this game somehow maxes that path which leads to the failure. I'm reading through the entire thread now, has anyone inspected the failed cards yet?

 read Igors Lab analysis. Crappy and outdated EVGA design
 
https://www.igorslab.de/e...n-caused-evga-problem/

Xavier Zepherious
currently every brand and type are affected
 
AMD
vega 64
rX500 series
and 6000 series
 
nvidia
1000 series
2000 series
3000 series
 
doesn't matter what AIB they are all affected
yes 3090 are the most affected because they have the highest power draw of any card
 
ive went thru all the post on all the forums
 
i dare say the cards with the most problems are the ones with the highest power draws
 
it could be a PSU overpower issue 
could also be that were driving the cards too hard and designers did not build in enough power protection to the GPU design or card design
this would be because engineers try to keep the cards cost down rather than build in excess protection
 
in any event this would be not the AIB fault but rather NVIDIA's and AMD fault for not requiring higher OCP standards and limits
 
ps: however realize this will drive the price of future card higher due to more protection measures  or higher quality parts  or more of them to be used
 

 
can we get some quotes about this other cards you found all over the internet. Not doubting you, but all points to a specific card, AIB and model problem.

Outofstock4ever

Xavier Zepherious
currently every brand and type are affected
 
AMD
vega 64
rX500 series
and 6000 series
 
nvidia
1000 series
2000 series
3000 series
 
doesn't matter what AIB they are all affected
yes 3090 are the most affected because they have the highest power draw of any card
 
ive went thru all the post on all the forums
 
i dare say the cards with the most problems are the ones with the highest power draws
 
it could be a PSU overpower issue 
could also be that were driving the cards too hard and designers did not build in enough power protection to the GPU design or card design
this would be because engineers try to keep the cards cost down rather than build in excess protection
 
in any event this would be not the AIB fault but rather NVIDIA's and AMD fault for not requiring higher OCP standards and limits
 
ps: however realize this will drive the price of future card higher due to more protection measures  or higher quality parts  or more of them to be used
 

 
can we get some quotes about this other cards you found all over the internet. Not doubting you, but all points to a specific card, AIB and model problem.

Well, JayzTwoCents posted on Twitter that he’s received word from followers that the following cards have all experienced this problem:

RX590
6800
6800XT
6900XT
3080Ti
3090

So, it doesn’t seem to be limited to a specific AIB or model GPU.

 
 

Outofstock4ever

ObscureEmpyre

Outofstock4ever

Xavier Zepherious
currently every brand and type are affected

AMD
vega 64
rX500 series
and 6000 series

nvidia
1000 series
2000 series
3000 series

doesn't matter what AIB they are all affected
yes 3090 are the most affected because they have the highest power draw of any card

ive went thru all the post on all the forums

i dare say the cards with the most problems are the ones with the highest power draws

it could be a PSU overpower issue 
could also be that were driving the cards too hard and designers did not build in enough power protection to the GPU design or card design
this would be because engineers try to keep the cards cost down rather than build in excess protection

in any event this would be not the AIB fault but rather NVIDIA's and AMD fault for not requiring higher OCP standards and limits

ps: however realize this will drive the price of future card higher due to more protection measures  or higher quality parts  or more of them to be used

can we get some quotes about this other cards you found all over the internet. Not doubting you, but all points to a specific card, AIB and model problem.

Well, JayzTwoCents posted on Twitter that he’s received word from followers that the following cards have all experienced this problem:

RX590
6800
6800XT
6900XT
3080Ti
3090

So, it doesn’t seem to be limited to a specific AIB or model GPU.

a bit biased and out of his league, also twitter. No mention to other cards failing anywhere especially amd cards. Zero. This seems to be a distraction to the real problem the ftw3 3090

Talk about confirmation bias.

Outofstock4ever

ObscureEmpyre

Outofstock4ever

Xavier Zepherious
currently every brand and type are affected

AMD
vega 64
rX500 series
and 6000 series

nvidia
1000 series
2000 series
3000 series

doesn't matter what AIB they are all affected
yes 3090 are the most affected because they have the highest power draw of any card

ive went thru all the post on all the forums

i dare say the cards with the most problems are the ones with the highest power draws

it could be a PSU overpower issue 
could also be that were driving the cards too hard and designers did not build in enough power protection to the GPU design or card design
this would be because engineers try to keep the cards cost down rather than build in excess protection

in any event this would be not the AIB fault but rather NVIDIA's and AMD fault for not requiring higher OCP standards and limits

ps: however realize this will drive the price of future card higher due to more protection measures  or higher quality parts  or more of them to be used

can we get some quotes about this other cards you found all over the internet. Not doubting you, but all points to a specific card, AIB and model problem.

Well, JayzTwoCents posted on Twitter that he’s received word from followers that the following cards have all experienced this problem:

RX590
6800
6800XT
6900XT
3080Ti
3090

So, it doesn’t seem to be limited to a specific AIB or model GPU.

a bit biased and out of his league, also twitter. No mention to other cards failing anywhere especially amd cards. Zero. This seems to be a distraction to the real problem the ftw3 3090

The failure rate would be a far more valid indicator of the issue - it's fair to say that NW stresses cards more than it should.
 
It's fair to say that all cards are more likely to fail under high stress.
 
Yet if 5% of EVGA 3090s fail but 1% of other cards under the same conditions - it's still fair to say that it's an EVGA 3090 issue.
 
No matter how you look at it, EVGA 3090 FTW3 is over represented in the failures.

There is always the fact that we don't know how many are actually true and just bandwagon on the interwebs. Some of the reports I see on Steam are just ridiculous.