z1nonly
My 3080 Ti is a 0.1.

Do *all* 3080 Ti's have new VRM's? (compared to the original 3080's and 3090's)

 
3080ti v0.1 =/= 3090 v0.1. The 3080ti came out after the 3090's were already at v1.0 and they have different VRMs than the 3090 did at v0.1. The PCBs are not the same between the two SKUs. I've seen the pictures to show that, and I've actually disassembled mine to put a hybrid kit on it and was able to confirm what chips were on there. I don't think there's been a revision to the 3080ti board. I don't have a 3090 though and I've only seen pictures of the v0.1 of those.

B0baganoosh

z1nonly
My 3080 Ti is a 0.1.

Do *all* 3080 Ti's have new VRM's? (compared to the original 3080's and 3090's)

 
3080ti v0.1 =/= 3090 v0.1. The 3080ti came out after the 3090's were already at v1.0 and they have different VRMs than the 3090 did at v0.1. The PCBs are not the same between the two SKUs. I've seen the pictures to show that, and I've actually disassembled mine to put a hybrid kit on it and was able to confirm what chips were on there. I don't think there's been a revision to the 3080ti board. I don't have a 3090 though and I've only seen pictures of the v0.1 of those.

Good info. Thanks.

Im having the black screen fans maxing issue when playing New World on my EVGA RTX3080 FTW3 Ultra. I have managed to cut the power and thus, I think, I have saved the GPU. This started happening to me after more than 150 hours in to the game. Then all of a sudden when one day I was booting the game and was just loading into the server this happened the first time. I cut power and rebooted, and then launched the game again thinking its was only an abnormality. It then worked fine again, only for it to happen again after about 45 minutes in the game. After that attempt I havent had the guts to start the game again, and this is the only game where this seemingly happens.
 
So my question is, what is the way to go here? Do I contact my retailer and demand an RMA or do I wait for a fix? Im really in the blue here not knowing what to do. I love the game and it would be a bummer to have to let it down, but I guess I could survive that. But I wouldn´t want to do that only for this happen later down the line in some other game.
 
Thanks for any reply to this, as i really don´t know how to go about this situation.

Bastinen
Im having the black screen fans maxing issue when playing New World on my EVGA RTX3080 FTW3 Ultra. I have managed to cut the power and thus, I think, I have saved the GPU. This started happening to me after more than 150 hours in to the game. Then all of a sudden when one day I was booting the game and was just loading into the server this happened the first time. I cut power and rebooted, and then launched the game again thinking its was only an abnormality. It then worked fine again, only for it to happen again after about 45 minutes in the game. After that attempt I havent had the guts to start the game again, and this is the only game where this seemingly happens.
 
So my question is, what is the way to go here? Do I contact my retailer and demand an RMA or do I wait for a fix? Im really in the blue here not knowing what to do. I love the game and it would be a bummer to have to let it down, but I guess I could survive that. But I wouldn´t want to do that only for this happen later down the line in some other game.
 
Thanks for any reply to this, as i really don´t know how to go about this situation.

There are a couple approaches here. I would bet that whatever the card uses to protect itself in the event that's triggering the safety shutdown (black screens, fans 100%) can only do it so many times. It might start triggering sooner, or eventually fail, which would result in a dead VRM and a blown fuse (ala the card BuildZoid evaluated). You can try reducing the power limit in Precision X1 to see if that buys you more time. You can keep hitting it with New World until it doesn't come back. You can also contact EVGA support and tell them about the problem to see if they'll give you an RMA before it actually pops. I'd say those are your options unless New World gets a patch that stops itself from doing...whatever it is that it does to cause the spiking behavior. I kind of feel like they're not going to do that because I don't think anybody is putting pressure on them. Folks like EVGA are RMAing cards, so Amazon will probably hope that just keeps happening.

B0baganoosh

Bastinen
Im having the black screen fans maxing issue when playing New World on my EVGA RTX3080 FTW3 Ultra. I have managed to cut the power and thus, I think, I have saved the GPU. This started happening to me after more than 150 hours in to the game. Then all of a sudden when one day I was booting the game and was just loading into the server this happened the first time. I cut power and rebooted, and then launched the game again thinking its was only an abnormality. It then worked fine again, only for it to happen again after about 45 minutes in the game. After that attempt I havent had the guts to start the game again, and this is the only game where this seemingly happens.
 
So my question is, what is the way to go here? Do I contact my retailer and demand an RMA or do I wait for a fix? Im really in the blue here not knowing what to do. I love the game and it would be a bummer to have to let it down, but I guess I could survive that. But I wouldn´t want to do that only for this happen later down the line in some other game.
 
Thanks for any reply to this, as i really don´t know how to go about this situation.

There are a couple approaches here. I would bet that whatever the card uses to protect itself in the event that's triggering the safety shutdown (black screens, fans 100%) can only do it so many times. It might start triggering sooner, or eventually fail, which would result in a dead VRM and a blown fuse (ala the card BuildZoid evaluated). You can try reducing the power limit in Precision X1 to see if that buys you more time. You can keep hitting it with New World until it doesn't come back. You can also contact EVGA support and tell them about the problem to see if they'll give you an RMA before it actually pops. I'd say those are your options unless New World gets a patch that stops itself from doing...whatever it is that it does to cause the spiking behavior. I kind of feel like they're not going to do that because I don't think anybody is putting pressure on them. Folks like EVGA are RMAing cards, so Amazon will probably hope that just keeps happening.

I think of myself as to casual (noobish) when it comes to adjusting powerlimits and so forth. So I guess my way forward will be to first pushing this to my retailer and see what course of action they would take while in the meantime just leave New World alone for time being and hope for a patch. I noticed a "semi-official" statement at the New world forum that they are still activly pursuing a solution to this, but as they at this point dont have a clue as to what causes this it will probably take some time.

B0baganoosh

pavan407
I've been desperate to play this game because of my love for MMORPGs. Although I have a 3080 Ti FTW3, as a safety measure I ran New World on my old rig with an ASUS GTX 780 Ti / 4790k (No overclock on both) @ 1080p 60FPS until there was more info available on the reliability with New World and 3080 Ti's. After 6 hours of gameplay with the lowest settings and 60 FPS cap in NVIDIA control panel, my PC blackscreened and GPU fans went jet mode. Had to force restart but luckily my 780 Ti is fine (I hope).
 
Coincidence? In my ~7 years of ownership of this card, not once has it blackscreened on any game. Haven't touched the game since let alone with my 3080 Ti FTW3.

This right here is why I put some blame on New World. There are lots of these reports, from all sorts of cards. Most cards will protect themselves (at least several times) and go into protect mode without dying. This is what is supposed to happen if some abnormal load causes an overcurrent on the gpu. The main issue I see with the 3090's is that they don't always do this correctly and end up dying. That's why I put most of the blame with the hardware.

The fact that New World can cause all sorts of cards from several years of different generations to go into protect mode means it is doing something wrong. It needs a patch to prevent the abnormal overstress it regularly puts on cards. That said, if your card does not protect itself and actually bricks from this event, it is a hardware fault and should be RMA'd, with full coverage of the card manufacturer. 

From Buildzoid latest video on the subject he believes it might be the undervolt protection triggering and not the over current protection simply because its damn near impossible to trigger the over current on the UPI 9511R controllers. I've RMA'd my first 3090 3 weeks ago due to 3 black screens/100% fans all from New World.
 
I'm being dumb and started playing it again but this time with a 70% power limit/120fps cap and 1080p windowed(was 1440p fullscreen). From the 3 hours I played last night my avg watts was 170 with a spike to 225w.

@ZoranC, I know I was a bit wordy, but I think you misinterpreted what I was saying. I also said that if the card dies, it is a hardware issue first and foremost.

There are reports of all kinds of cards of various brands and generations that go into protection mode while playing New World (not die, but go into protection mode... Well... Some die). This shouldn't happen in a game. It's not normal and to act like New World is doing nothing wrong is ridiculous. Is it Amazon's fault the cards actually die? No. But there's no good reason for what their game is doing. It is creating an overstress that they don't even understand. There are plenty of cards, like the 780ti above that are protecting themselves correctly, and that's what should happen in the case of an overstress like they are experiencing. They should shut down and protect themselves so they don't die. The cards that fail instead have a design fault.

That said, why is this particular game causing a situation on many cards where they have to protect themselves from burning up? It's not even a graphically demanding game. This shouldn't happen. There's no good reason for it. I agree 100% that the cards shouldn't "jump off a cliff". But using your example, why is the game asking them to? Or to use the other example, why is it bypassing the physical movement of the pedal to hit the digital pedals (let's get modern here) and asking the engine for 0% throttle, then 200% then 0% then 200% in rapid succession until the computer gets confused and puts the car in "limp home" mode? Sorry, the car example is not really relevant considering the vast differences in architecture and how many variables there are in the circuit topologies and software, firmware, etc., but I tried to use it seeing how you brought it up.

In terms of the 3090 FTW Rev 0.1s dying, I know more than one game has caused them to fail. That's why I said the primary fault for those was with EVGA and Nvidia for the design decision and guidance/approval.

The car thing just isn't working here. A graphics card has more than just a gas pedal, it isn't as simple as "go full throttle". Plenty of cards are dying while not running anywhere near maximum power.

There are two things I've been referring to here, one is v0.1 3090FTW cards. These were dying in other applications too. Clearly a hardware issue. The second topic is the fact that New World is putting various cards into a safety shutdown. Cards that don't have that issue in any other application. So New World is exacerbating a hardware issue already present with 3090FTW v0.1 cards, but also overstressing other cards for no legitimate purpose (its not intended by the developer at all).

We agree on the first point. You then think my second point is at odds with the first, but they are not mutually exclusive. New World is more stressful on a gpu than a "stress test" is without even fully utilizing the graphics card. It is creating a special situation that the cards have to shut down and crank the fans in order to save themselves. Why should they not try and fix whatever is causing that? It's not just "throttle to 100%" it's more like they plug in a computer tuner when they get in and it doesn't set things up properly so some cars die, even if the driver is just cruising down the road going 30mph at 1800 rpm (just trying to use the car comparison you won't let go of). The car(d) shouldn't die, yes, but why wouldn't you want them to also fix the crappy tuning software?

and the answer is no i still got another 0.1 rev after asking for a 1.0

it will probbaly blow again, after that probably will request complete refund cause 3090's kinda suck bruh


the difference between the one i got this time and the last time is this one has a QC sticker on it.

ZoranC

B0baganoosh
The car thing just isn't working here. A graphics card has more than just a gas pedal, it isn't as simple as "go full throttle". Plenty of cards are dying while not running anywhere near maximum power.

There are two things I've been referring to here, one is v0.1 3090FTW cards. These were dying in other applications too. Clearly a hardware issue. The second topic is the fact that New World is putting various cards into a safety shutdown. Cards that don't have that issue in any other application. So New World is exacerbating a hardware issue already present with 3090FTW v0.1 cards, but also overstressing other cards for no legitimate purpose (its not intended by the developer at all).

We agree on the first point. You then think my second point is at odds with the first, but they are not mutually exclusive. New World is more stressful on a gpu than a "stress test" is without even fully utilizing the graphics card. It is creating a special situation that the cards have to shut down and crank the fans in order to save themselves. Why should they not try and fix whatever is causing that? It's not just "throttle to 100%" it's more like they plug in a computer tuner when they get in and it doesn't set things up properly so some cars die, even if the driver is just cruising down the road going 30mph at 1800 rpm (just trying to use the car comparison you won't let go of). The car(d) shouldn't die, yes, but why wouldn't you want them to also fix the crappy tuning software?

 
Unfortunately I don’t have enough time to respond to your post in detail so I will briefly point out few things which I believe have been already pointed out:
 
1. From what I am reading it seems that you are incorrect in your statement that New World is only application triggering safety shutdowns.
 
2. You are incorrect in believing it is just New World (or any other software) and video card that are involved in final result. When piece of graphics software wants something rendered it (roughly) calls graphic library (responsibility of Microsoft), graphic library interacts with card’s driver (responsibility of Nvidia), driver interacts with card (responsibility of Nvidia and its partner), and also card’s firmware is involved (again responsibility of Nvidia and its partner). So New World or any other piece of software could be making what seemed a perfectly legitimate call to developer and accidentally created perfect storm condition. Should they all be looking into it together? Yes. But way I see it New World here is a victim of circumstance. No modern end user graphic software that I know of interacts directly with hardware.
 
Also, your “crappy tuning software” analogy is missing one huge point. In this case “tuning software” would be card’s driver and/or firmware. Which is again not responsibility of New World but Nvidia and its partners. New World is just “person” behind the wheel that is trying to drive that car and has requested it accelerates.

*sigh*...again, that's not what I said. 
 
To expand on what I did say and get away from the car analogies that don't really work here at all, I'll add this:
New World should be working with Nvidia, AMD, and Microsoft (and maybe even others, I don't know) to try and figure out why their software is putting so many cards into a safety shutdown condition. Should it happen? no. It is happening though. It is happening to 3090s, 3080s, 780tis, 1070s, AMD cards, just about every card out there can have this happen and there are reports of it happening. Why is that happening? Should it be fixed with a game update or a driver update? Maybe both. Furmark used to kill cards, so Nvidia came up with a driver update that detected Furmark and throttled the card. So maybe Nvidia and AMD need to throttle the cards while people play New World. I don't feel like that's the best answer, but it might be part of the solution. The problem with that idea is that older cards are not supported by the new drivers anymore. So New World might still cause problems for those cards. Your suggestion seems to be that "every card that exists has a design fault" (not an actual quote, just a paraphrase). Doesn't that sounds much less reasonable than "maybe New World is doing something wrong"?
 
New World is not the only application that has ever caused the shutdowns. I didn't say it was. I did say that it is causing shutdowns on all kinds of graphics cards. This needs to be addressed, preferably by a patch to their game and working with Nvidia, AMD, and probably Intel soon to get updated drivers out there that limits the bad behavior causing problems.
 
It is not Amazon's fault that cards died. It is their fault (and maybe Microsoft, AMD, and Nvidia's fault too, which I did not indicate otherwise by the way) that their software is causing this "perfect storm" or bad situation. So they should be involved in making that not happen anymore. There's no good reason for it to happen. You literally reiterated my point and then said I was wrong lol. I think you read it with a very binary logic in that if I said New World/Amazon had some responsibility for causing the bad situation (shutdowns) that I meant they alone own all responsibility and that's not what I said. It's not 0 or 1. They share a piece of the pie. They have some responsibility for their game causing shutdowns in so many graphics cards.
 
All of that said, all of my theories here are 100% conjecture based on the evidence we have and data provided by users and industry experts that have spoken about it. We don't have any confirmation from Nvidia, EVGA, Amazon, or anybody else that they know exactly why people's graphics cards have shutdowns in New World, or why a bunch of 3090's (and some other cards) are dying when they try to play this particular game (yes, I know that the v0.1 3090 FTW3 also had a tendency to die in Halo:MC, LoL, and a couple other older games). Until there's confirmation or enough data is collected to verify independently, we won't know for sure. 
 
We also know that some v0.1 3090 FTW3s have been operating completely fine for people in every application tested. So we can't even say with full confidence that every single one of them is bad. It seems there was enough concern about their design that EVGA revised the board and VRM circuit, but that doesn't mean that all of the v0.1's will die. 

ZoranC
I'm not aware of a single report of it happening to AMD cards. If you are I would appreciate links to them, please, so I can inform myself better.

Sure thing:
6800XT (if you look in the rest of this thread, there are 3080tis, 3080s, a 2070, and 1070 from various brands with shutdowns and some deaths)
6800
6900XT (crashing unless capping FPS)
6900XT
6800XT (Also a 3070 and 3080s in this thread)
6900XT
6800
6900XT
 
There are other sites too, but that gives you the idea. It is notable that most of them just have a crashing issue, some hard crashing requiring some downtime and unplugging/re-plugging things to get it back, others just black screens and/or driver crashes. Only a few of these incidents I found in a quick search actually resulted in a bricked card. So as I had said before, the cards should protect themselves (and most do), but this game seems to be very unstable and causes issues for a lot of people. I mean, just search "game crashing" on there. It's definitely got issues.

ZoranC
Which further confirms need for all vendors involved in that chain to work together in the best interest of their consumers.

I completely agree!

nightskycode
Does anyone know when the RMA arrives if I need to sign for it? If so I may have to have it held at the post office for pick up as I'm always at work. It says 24G-P5-3982-B2 was the part number shipped, differnat than my other 3090, so I think it may be a Rev 1.0. Either way, I never played new world, never will. It crashes for me on halo and my 3D modeling software I use for work. Card never died, it did a good job at revving the fans up and saving itself. Though, I had enough of the lost work.

Mine did not require a signature.

bill1024
No signature for now with UPS because of covid. 

That's good to hear, though I may change the delivery option on UPS's website to pickup. A 3090 is not something you want some random highschool kid on a walk to snag off your porch.It's never happened to me, though if it did happen, it would be this package lol

As a 3090 FTW3 Red-lips owner circa November 2020 SN, I can confirm here that an FPS cap in Rivatuner and Nvidia Settings along with decreasing your power limit to 75% is not completely sufficient to protect the card. I've played approx 100-150 hours in new world with it using those settings so I was starting to believe it was a safe solution, but when I teleported from one city to another it crashed the computer and bricked the card like so many others. Now the wait begins! Thanks everyone for the info on this thread it's been very informative.