Amazon's mmo New World is bricking 3090 gpus

[spartaman64]

Quote

Reports on the New World subreddit indicate that trying to launch New World while using an RTX 3090 will sometimes cause it to completely fry, forcing those impacted to contact manufacturers for refunds or replacements. The issue seems to be affecting EVGA RTX 3090 cards more than other GPUs, though it's worth noting that people who use other versions of the 3090 have reported problems with overheating and dangerous power spikes.

Quote

The New World team hasn't yet officially recognized the issue nor offered advice on avoiding it or any fixes. Many experiencing problems are saying that the lobby is causing the issues, with insane FPS spikes for extended periods of time causing GPU temperatures to get out of hand and then the hardware problems begin. But others are also reporting that they're getting into a game and playing for a while before things start. Usual tricks like V-Sync or FPS caps don't seem to be fixing it, either.

 

edit new world's response: 

 

source: https://www.windowscentral.com/amazons-new-world-bricking-rtx-3090-graphics-cards 

 

I've seen speculation that it might be the memory chips which do tend to run very hot and with this extra load maybe they are frying themselves. But whatever the reason I'm going to avoid new world until this has been addressed since I don't want it to fry my 3080 which is hard to replace at the moment. 

[tkitch]

upside?  (pretty much) All cards should still be under warranty?  

 

Still sucks, and that's hella poor design somewhere.

[Spectrox_]

3 minutes ago, tkitch said:

upside?  (pretty much) All cards should still be under warranty?  

 

Still sucks, and that's hella poor design somewhere.

Where are they gonna get the replacements in a shortage?

[tkitch]

1 minute ago, Spectrox_ said:

Where are they gonna get the replacements in a shortage?

if you think EVGA and other OEMs don't keep spares around, you're mistaken.

[JLO64]

2 minutes ago, tkitch said:

if you think EVGA and other OEMs don't keep spares around, you're mistaken.

I've heard reports that individuals who have had to get RMAs for 30-series cards couldn't be given a replacement due to the shortages and were given cash only refunds. Granted, that was several months ago so hopefully the situation has improved on that front.

[Just that Mario]

I am surprised that some game can even cause this. Aren't graphics cards supposed to have multiple failsafes for such overheating, etc? I am not sure to be impressed or worried over this issue.

[Skiiwee29]

3 minutes ago, JLO64 said:

I've heard reports that individuals who have had to get RMAs for 30-series cards couldn't be given a replacement due to the shortages and were given cash only refunds. Granted, that was several months ago so hopefully the situation has improved on that front.

Thats not so much that they didn't have replacements, but that they could make more money off refunding and continue to sell them at the jacked up scalper prices and make the money while they could. 

[rcmaehl]

Big Yikes. As @JLO64 said,there hasn't really been replacements/refurbs lately. Hopefully, if refunded, they'll get current asking prices instead of MSRP otherwise they'll be without a GPU for a while.

 

[Helpful Tech Witch]

9 minutes ago, tkitch said:

if you think EVGA and other OEMs don't keep spares around, you're mistaken.

Yeah, people who’ve rmaed their Gpus recently have gotten full refunds instead of new cards

[Avocado Diaboli]

Maybe this is a dumb take, and yeah, the game is probably terribly optimized. But wouldn't a hardware failure like that indicate that this is more of a fault with the hardware manufacturer and Nvidia's specifications of these cards? I mean, sure, the devs should address the problem, but honestly, I feel like I shouldn't have to worry that some random piece of software that I run might be metaphorically exploding my PC.

[Fasterthannothing]

Yikes seems like some incompetent devs at Amazon. They already ruined their game I never heard of it before so I googled it and couldn't find anything except articles saying it's killing computers not a good PR situation.

[CarlBar]

13 minutes ago, Avocado Diaboli said:

Maybe this is a dumb take, and yeah, the game is probably terribly optimized. But wouldn't a hardware failure like that indicate that this is more of a fault with the hardware manufacturer and Nvidia's specifications of these cards? I mean, sure, the devs should address the problem, but honestly, I feel like I shouldn't have to worry that some random piece of software that I run might be metaphorically exploding my PC.

 

Sure but why is this the first thing to do it.I'd expect a serious hardware fault to have started showing itself in other things before this.

 

My best guess is the games asking the card to do things that are highly atypical, (probably because it's not considered a good way to do a particular thing). Another possibility is the protections on the card have a flaw that somthing about how this software works is accidentally bypassing or disabling.

[WereCat]

So Amazon made a new playable version of Furmark? 

[DildorTheDecent]

It's probably just killing the last few remaining FTW3s that weren't already RMA'D for the power delivery problem.

 

Although it wouldn't surprise me if the RMA GPUs are junk too.

[RejZoR]

How is this Amazon's fault, realistically? It's NVIDIA's responsibility that cards cannot make power spikes that burn out their VRM's, it's NVIDIA's responsibility that cards individual modules don't burn out when under extreme loads. It's entirely NVIDIA's responsibility and of vendors that use their GPU's.

 

I always rolled my eyes when they were complaining over FurMark damaging cards. It just shouldn't matter. Design better cooling, power delivery or safety systems to prevent this and not blame users.

 

These RTX 3090's dying seems more like it's connected with their garbage cooling of VRAM's that were problematic as is and tend to overheat under normal use or run at its very edge of thermal limit.

[wanderingfool2]

1 hour ago, CarlBar said:

(probably because it's not considered a good way to do a particular thing).

More like an overlooked thing by the hardware that isn't common...but uncommon doesn't mean it's bad.  An example being Intel's bug...divide 4,195,835 by 3,145,727 (on an old pentium) and you get the wrong result.  Similar things could be happening here, where it's hitting an obsure case that's causing issues.

 

Ultimately though, software should not be blamed for killing hardware like this...hardware should have the adequate safeguards in place to prevent bricking.  If it's because it's being stressed so much that the system is breaking, then I would be willing to guess that in a year or two those cards might have fried themselves anyways...this just is accelerating it.  This is more of a hardware design failure

[CerealExperimentsLain]

2 hours ago, tkitch said:

upside?  (pretty much) All cards should still be under warranty?  

 

Still sucks, and that's hella poor design somewhere.

"Your RTX 3000 RMA will be full filled in... 7 years."

[Helpful Tech Witch]

16 minutes ago, RejZoR said:

How is this Amazon's fault, realistically? It's NVIDIA's responsibility that cards cannot make power spikes that burn out their VRM's, it's NVIDIA's responsibility that cards individual modules don't burn out when under extreme loads. It's entirely NVIDIA's responsibility and of vendors that use their GPU's.

 

I always rolled my eyes when they were complaining over FurMark damaging cards. It just shouldn't matter. Design better cooling, power delivery or safety systems to prevent this and not blame users.

 

These RTX 3090's dying seems more like it's connected with their garbage cooling of VRAM's that were problematic as is and tend to overheat under normal use or run at its very edge of thermal limit.

its not nvidias fault if the aib makes a shoddy card that doesnt work, same for the memory

[WkdPaul]

As far as I've seen, the only 3090s that die from this are the EVGA 3090 FTW3, other 3090s have heat and throttling issues, but as far as I've seen haven't totally died like the FTW3.

[WereCat]

8 minutes ago, wkdpaul said:

As far as I've seen, the only 3090s that die from this are the EVGA 3090 FTW3, other 3090s have heat and throttling issues, but as far as I've seen haven't totally died like the FTW3.

I find it interesting how it's always FTW3 that has the issues. 

The 1000 series had issues with cooling, then they made the ICX series to fix it. 

Then their first batch of FTW3 hybrid on the 1000 series had a cooler issue that was bricking the cards. 

Now this... 

[PocketNerd]

26 minutes ago, RejZoR said:

How is this Amazon's fault, realistically? It's NVIDIA's responsibility that cards cannot make power spikes that burn out their VRM's, it's NVIDIA's responsibility that cards individual modules don't burn out when under extreme loads. It's entirely NVIDIA's responsibility and of vendors that use their GPU's.

 

I always rolled my eyes when they were complaining over FurMark damaging cards. It just shouldn't matter. Design better cooling, power delivery or safety systems to prevent this and not blame users.

 

These RTX 3090's dying seems more like it's connected with their garbage cooling of VRAM's that were problematic as is and tend to overheat under normal use or run at its very edge of thermal limit.

Someone doesn't know how GPUs are actually made.

 

It would be EVGA's fault for not adding enough cooling to their card(s), not Nvidia.

 

That said, I don't think this is a cooling issue...

[MageTank]

This is interesting that it's only limited to 3090's. I played the Alpha with my 2080 Ti and found my thermals to be quite tame compared to titles like Overwatch under extended loads (surprisingly, this game used a fair bit of my 5950X despite being an MMO, modern DX12 engines are neat).

 

Despite my disdain for the combat system and the inability to jump in the game (overall clunky feel), it performed really, really well for how beautiful it looked. I also recall quite a few people in the world chat talking about their system builds in the Alpha and I know a few mentioned having 3090's but none mentioned any thermal issues when the GM was doing performance tests in large areas. Something may have changed in the beta, but I haven't played since then as the game isn't my cup of tea.

[Middcore]

This is very hard to believe. If I had a dollar for every time I heard someone claim "this game killed my GPU!"...

 

And why would it only be 3090's and not lesser cards? And why were there no dead cards during the alpha a few months ago?

[WkdPaul]

10 minutes ago, Middcore said:

This is very hard to believe. If I had a dollar for every time I heard someone claim "this game killed my GPU!"

 

And why would it only be 3090's and not lesser cards? And why were there no dead cards during the beta a few months ago?

It's one specific model, the EVGA FTW3, it's not ALL 3090s.

 

There are multiple reports of this on twitter, reddit and even New Worlds' own forums, it's not just 2-3 people. Some even had their RMA 3090 FTW3 die also (this was already there during the alpha).

Edited July 21, 2021 by wkdpaul

[xg32]

some1 whose card was undervolted and survived reported 84C core and  115C memory, and so far only evga cards are consistently bricking, the ram temps suck but it's not the ram, signs point to high fps in the menu and likely triggering a specific component on the ftw3, similar to coilwhine, other hardware have reported higher temps regardless but no bricks.