Perpixel
In terms of visual damage, does anyone know what to look for on the board? I have some good pictures of my card before sending it back to EVGA. I was unable to see anything obviously wrong even after the pretty loud POP it did.

gsrcrxsi

Perpixel
In terms of visual damage, does anyone know what to look for on the board? I have some good pictures of my card before sending it back to EVGA. I was unable to see anything obviously wrong even after the pretty loud POP it did.

post the pictures.

Perpixel

gsrcrxsi

Perpixel
In terms of visual damage, does anyone know what to look for on the board? I have some good pictures of my card before sending it back to EVGA. I was unable to see anything obviously wrong even after the pretty loud POP it did.

post the pictures.

 
Here,

EVGA_FTW3 - Google Drive
 
{EDIT}
 
I assume the 10 amp is the one at the back without a clear picture.

gsrcrxsi
 
yes I believe its the one on the back close to the PCIe slot.
 
if you had a multimeter, you could have checked the fuses for continuity, but it sounds like you already sent this back to EVGA.
 
I don't see any obvious signs of components being damaged. but they could certainly pop internally without visible damage. also possible that the pop you heard was from another component in the PC (like the PSU) as OCP was tripped.

Perpixel

gsrcrxsi

Perpixel
In terms of visual damage, does anyone know what to look for on the board? I have some good pictures of my card before sending it back to EVGA. I was unable to see anything obviously wrong even after the pretty loud POP it did.

post the pictures.

 
Here,

EVGA_FTW3 - Google Drive
 
{EDIT}
 
I assume the 10 amp is the one at the back without a clear picture.

ComboSlicer
https://www.igorslab.de/en/evga-geforce-rtx-3080-rtx-3090-and-not-only-new-world-when-the-graphics-card-goes-amok-because-of-design-failures/
 
As per igorslab seems it is a faulty fan controller, I hope tjis can be fixed with a firmware update. Any meaningful update from evga so far?

ComboSlicer
https://www.igorslab.de/en/evga-geforce-rtx-3080-rtx-3090-and-not-only-new-world-when-the-graphics-card-goes-amok-because-of-design-failures/
 
As per igorslab seems it is a faulty fan controller, I hope tjis can be fixed with a firmware update. Any meaningful update from evga so far?

hallmark888
I have the FTW3 Rev 0.1 (Red lipped card, that occationally hit 82W PCI-E, avg around 77W) which I bought in November of 2020, I beta tested the game and the card was fully stable during the play session (with 100% and 119% PL). No black screen, Fan 100% or any of that jazz. If your card behaves this way it is most likely a dud. What's interesting is if card is at 119% PL it will constantly draw 475+ W, and will transiently hit about 505W power , which is not the case for most other games (Avg 440W in non-RTX games). I've only seen that power draw pattern in MSI Kombustor v4 GL-msi01. 

hallmark888
1830-1875MHz @ 0.937-0.981mV, in-game 99% usage.
 

hallmark888

kevinc313

hallmark888
1830-1875MHz @ 0.937-0.981mV, in-game 99% usage.
 

 
Just to be clear, you're pulling 500w on a XOC bios at the above frequency / vcore?
 
I agree that is like a Furmark load. Is it sustained?
 
If someone has a .950V undervolt applied with max PL and limited cooling, that could be a problem for their temps / vram temps.

Yes. That's using the XOC bios, running everything at 0 offset. So basically stock card with 119% PL. Card will remain steady at the clocks above. 
Other workloads that have similar pattern from my testing are MSI Kombustor v4 GL-msi01 (very similar load pattern) and second is Quake II RTX, but Quake draws like 20W less, never seen it come close to 500W, but maxes around 480W.
Furmark though, seems to throttle the card between 1400-1500MHz, which isn't the case for NW.

KeithW19

AJKagan
I am also using my main rig for WFA purposes which is why I didn't even think about chancing it when OCP kicked in. It happened once, and that was enough for me to uninstall. I knew that whatever their "fix" was, was clearly not an actual fix.
 
Not worth the risk to my daily 8-5, not over a $40 pre-order beta game. Not over ANY game tbh.  Thankfully I just recently bough this asus gaming laptop with good specs (Ryzen + 3060) so it can game reasonably well. Just not ideal to be gaming on a laptop.
 
 

But its not the game. In other words yes you can avoid New World and your card will continue to work. But its a ticking time bomb. Whose to say the next game released wont also have the same issue and kill your card? Do you just wait on every game before you play to see if it kills these cards and then if it does you never play it? Thats not a functional product. Check out my link in the post below yours

kevinc313
Here is an early report of a Gigabyte 3090 black screening immediately on the startup of the game, when there was an uncapped framerate in the menu.
 
https://forums.newworld.com/t/issues-with-evga-rtx-3090-ftw3-ultra/112757/39
 
Gigabyte cards have the worst VRAM thermal pads, GDDR6X junction temp will run like 20C hotter than an equivalent FTW3 at the same load and it is possible to peg 110C almost instantly on their cards.  It is conceivable that the game did that and crashed/protected the card.

Kylearan

kevinc313
Here is an early report of a Gigabyte 3090 black screening immediately on the startup of the game, when there was an uncapped framerate in the menu.
 
https://forums.newworld.com/t/issues-with-evga-rtx-3090-ftw3-ultra/112757/39
 
Gigabyte cards have the worst VRAM thermal pads, GDDR6X junction temp will run like 20C hotter than an equivalent FTW3 at the same load and it is possible to peg 110C almost instantly on their cards.  It is conceivable that the game did that and crashed/protected the card.

It's worth noting that the Gigabyte card also has fuses.
Seems like all the cards that have reported failures all seem to have fuses (eVGA except Kingpin, Gigabyte and MSI).
Other AIB cards have crashed but no magic smoke.

gsrcrxsi

GlaucomaPredator

gsrcrxsi

ComboSlicer
https://www.igorslab.de/en/evga-geforce-rtx-3080-rtx-3090-and-not-only-new-world-when-the-graphics-card-goes-amok-because-of-design-failures/

As per igorslab seems it is a faulty fan controller, I hope tjis can be fixed with a firmware update. Any meaningful update from evga so far?

Nah this is a wack theory.

Why?

because it's only an observation of some phenomenon with no real conclusion or insight into HOW this phenomenon could possibly cause the issues with dead GPUs.
 
the whole article is "AHA! look at this big RPM number! Smoking gun!" and no real details. if it's a problem then explain how this could possibly kill a GPU.
 
It can't. it's nonsensical. as another user posted, in order to attempt to drive (if that's the implication here) a fan at higher than max RPM, would require more than 12V. and with no step UP voltage converters on the board, where would such voltage even come from?
 
looks like this specific card (the unit in question, not the whole line) has some problem with the fan control, but it's not dead, and wasn't killed by the game even though it clearly exhibits this strangeness, so obviously not the root cause. and this issue with fan control isn't as widespread as he claims. I've never seen it mentioned on these forums though he claims there's a mountain of posts about it.
 
IMO his card has some problem with the fan controller that's spitting out faulty RPM readings. and that's totally separate from what's happening with the New World game.