AquinusThere have been plenty of GPUs in the past that have burnt themselves up. This isn't a new phenomenon. It's also not the game's fault, it's usually the device or drivers.

Agreed. This generation I don’t recall widespread reports of any cards faulting upon the release of a game.

New World releases in closed beta. Widespread reports of GPU faults flood in, with EVGA being spot lighted. Amazon Games has responded to this and has reported a bug upon reports, especially in 4K, but has not specifically claimed fault. EVGA has responded by saying they will uphold warranties.

R-T-BDo closed betas usually wreck your PC physically?

Error prompt or burning PC, its all the same man. Gosh. Get with the outrage culture program already!

Caring1No Jay, it's how the drivers handle the game.

Jay doesnt need to be right, protruded eyeballs and all caps generate enough clicks for his livelihood.

RavenasAgreed. This generation I don’t recall widespread reports of any cards faulting upon the release of a game.

New World releases in closed beta. Widespread reports of GPU faults flood in, with EVGA being spot lighted. Amazon Games has responded to this and has reported a bug upon reports, especially in 4K, but has not specifically claimed fault. EVGA has responded by saying they will uphold warranties.

I would certainly not discount the thought thosr shite Ampere GPUs are already hanging together with duct tape and other bandaids. The problems they have had are all attributable to temps and current. They are power hogs, all of them. Spiky, and pretty unrefined. The 3090 crashing is not new to nvidia top end. 2080ti space invaders... its clear they walk a fine line , these products. Enough writings on the wall most notably the weird ti releases and VRAM capacities and the increased TDP across the stack.

Tenfoot poles still apply even if you can buy them IMHO.

A game won't make a GPU exceed it's full load cooling and power consumption poorly designed hardware that was pushed beyond it's breaking point however will. The RTX 3080/3090 is the poster child of hardware being really over aggressively pushed to it's knees to a questionable degree. Nvidia literally made a new power connection because the standard one that had been around for decades simply wasn't enough for it's power hungry FPS crown it was after and it needed to push TDP a lot higher better offset the RX series strong rasterization performance that was and is arguably much more compelling to many.

Nvidia was already pushing things heavily AIB's took it further and then overclocking enthusiasts got them and we're like we'll I test my luck too what's the worst that can happen my house burns down when I'm AFK in the New World game menu?

InVasManiThe default TDP is as high as my Seasonic PSU's entire wattage. Suffice to say I won't be getting a RTX3080 now or later.

www.techpowerup.com/forums/threads/evga-releases-optional-bios-update-for-rtx-3080-ftw3-ultra-xoc-450-w.273372/#post-4369689

InVasManiCuda core count alone doesn't reflect a linear gain in performance. They still have power and TDP to take into account and more Cuda cores will result in more power and heat though how much depends on the node design as well.

BTW did we ever a get a clear answer to the cause of the space invaders fiasco on previous RTX series cards or was just swept under the rag and forgotten!!? There was plenty of speculation about it, but was any legitimate consensus by Nvidia or partners disclosed in the end!!?

Vayra86Jay doesnt need to be right, protruded eyeballs and all caps generate enough clicks for his livelihood.

I don't believe Jay is over concerned if he is correct or not, but would certainly prefer to be. If I can say one thing about Jay he is pretty genuine about his tinkering with hardware and willing to make personal investigations for the spirit of seeing how hardware behaves. He's knowledgeable enough, but not a absolute expert in all areas and knows as much and willing to concede he's wrong as well. Linus about MT/s not so much.

RavenasAgreed. This generation I don’t recall widespread reports of any cards faulting upon the release of a game.

New World releases in closed beta. Widespread reports of GPU faults flood in, with EVGA being spot lighted. Amazon Games has responded to this and has reported a bug upon reports, especially in 4K, but has not specifically claimed fault. EVGA has responded by saying they will uphold warranties.

would you say the same about furmark? it's furmark's fault my card's shitty VRM config blew up! furmark made my GPU consume 200% power and blow up!
furmark HAKECED my GPU! ban he! now!

Vayra86Error prompt or burning PC, its all the same man. Gosh. Get with the outrage culture program already!



Jay doesnt need to be right, protruded eyeballs and all caps generate enough clicks for his livelihood.



I would certainly not discount the thought thosr shite Ampere GPUs are already hanging together with duct tape and other bandaids. The problems they have had are all attributable to temps and current. They are power hogs, all of them. Spiky, and pretty unrefined. The 3090 crashing is not new to nvidia top end. 2080ti space invaders... its clear they walk a fine line , these products. Enough writings on the wall most notably the weird ti releases and VRAM capacities and the increased TDP across the stack.

Tenfoot poles still apply even if you can buy them IMHO.

I dont think you should compare this issue to space invaders.

InVasManiI don't believe Jay is over concerned if he is correct or not, but would certainly prefer to be. If I can say one thing about Jay he is pretty genuine about his tinkering with hardware and willing to make personal investigations for the spirit of seeing how hardware behaves. He's knowledgeable enough, but not a absolute expert in all areas and knows as much and willing to concede he's wrong as well. Linus about MT/s not so much.

Its TV. Its an act. Come on. Jay is what he wants us to see.

AuerI dont think you should compare this issue to space invaders.

The cause of the issue may not be the same, but the general quality of the top-end GPUs in Nvidia's stack has a specific note to it. Every time, and logically so, its the hottest head in the stack, walking the finest line and with a relatively low OC potential. This is not every gen - but the 780ti was a similar bag of beans. Hot, low OC potential. Basically what you got out the box was what you got. And stability? Not fantastic. Its very noticeable you're dealing with a bigger chip than any x104 unit they produce.

They áre cut down, less perfect Titans (ánd produced on a node they didn't quite prefer). Lets not fool ourselves here just because it commands 1K price tags or more. As the technology gets more refined, so does power circuitry and Nvidia has been fiddling a lot with that - with varying success. Pascal was great, Maxwell was a step forward, and Kepler was a step up from Fermi, but at some point something's gonna give. Turing and Ampere are that moment, if you ask me.

Vayra86Its TV. Its an act. Come on. Jay is what he wants us to see.


The cause of the issue may not be the same, but the general quality of the top-end GPUs in Nvidia's stack has a specific note to it. Every time, and logically so, its the hottest head in the stack, walking the finest line and with a relatively low OC potential. This is not every gen - but the 780ti was a similar bag of beans. Hot, low OC potential. Basically what you got out the box was what you got. And stability? Not fantastic. Its very noticeable you're dealing with a bigger chip than any x104 unit they produce.

They áre cut down, less perfect Titans (ánd produced on a node they didn't quite prefer). Lets not fool ourselves here just because it commands 1K price tags or more. As the technology gets more refined, so does power circuitry and Nvidia has been fiddling a lot with that - with varying success. Pascal was great, Maxwell was a step forward, and Kepler was a step up from Fermi, but at some point something's gonna give. Turing and Ampere are that moment, if you ask me.

And yet, Space Invaders affected but a fraction of real world users and I suspect this latest issue will be even more shortlived.
EVGA has been sending out new cards to users with issues since yesterday.
You seem to think this is an ongoing issue that we will keep hearing more about, well maybe not.

AuerAnd yet, Space Invaders affected but a fraction of real world users and I suspect this latest issue will be even more shortlived.
EVGA has been sending out new cards to users with issues since yesterday.
You seem to think this is an ongoing issue that we will keep hearing more about, well maybe not.

Not at all? That's you jumping to my conclusions ;) I'm saying, these cards walk a fine line, so bad batches are too hard to spot and actually end up in the wild; and games can tweak a menu in such a way it causes them to burn to a crisp.

Unpatched and yet older GPUs with slightly less performance are fine? Mmm. Interesting. 500 FPS is ok, 1000 is not? Mmmm I see.

It makes you wonder 'what's next' for these cards. You don't know. And that's pretty serious for halo product.

Its also not exclusive to Nvidia or its top end cards, mind. Its just a fact every GPU in every gen's stack is a different one with different cost cutting, die cutting measures, and systems that are implemented in different ways. And there is definitely a bit of variation between them, some being great and others not quite so.

Another item of note is what you mention: "EVGA". A company notorious for problems with self designed cards/coolers.

Vayra86Not at all? That's you jumping to my conclusions ;) I'm saying, these cards walk a fine line, so bad batches are too hard to spot and actually end up in the wild; and games can tweak a menu in such a way it causes them to burn to a crisp.

Unpatched and yet older GPUs with slightly less performance are fine? Mmm. Interesting.

It makes you wonder 'what's next' for these cards. You don't know. And that's pretty serious for halo product.

Its also not exclusive to Nvidia or its top end cards, mind.

Lets just all tip our hats to EVGA that provides us with bleeding edge cards, Insane available 500W+ Bioses and then accepts RMA's when we fry shit.
I believe there's stuff out there for the Camry drivers too, it's all good.

AuerLets just all tip our hats to EVGA that provides us with bleeding edge cards, Insane available 500W+ Bioses and then accepts RMA's when we fry shit.
I believe there's stuff out there for the Camry drivers too, it's all good.

Bleeding edge? They offer the same stuff every other AIB does. And their warranty is not just service, its an absolute necessity given the fuckups of recent years/gens. Our thoughts are different here. That's fine.

Vayra86Bleeding edge? They offer the same stuff every other AIB does. And their warranty is not just service, its an absolute necessity given the fuckups of recent years/gens. Our thoughts are different here. That's fine.

Yeah you know, glass half full and all that ..

Vayra86Its TV. Its an act. Come on. Jay is what he wants us to see.

You just described the entire internet basically.

JAB CreationsThe problem is that unless there is a configuration file that you already know about and are absolutely certain how to turn V-Sync on then you run the risk of a 7,000+FPS black screen between you and turning it on in the game's menus. Most people aren't going to know that until either they come across the news (presuming they read or watch tech news) or their card gets killed.

Nvidia control panel allows for limiting to a max fps value prior to first launch. So in the worst case, that would be a solution. Not that this subtracts anything from the unacceptability of such an issue to begin with.

YtterstaNvidia control panel allows for limiting to a max fps value prior to first launch. So in the worst case, that would be a solution. Not that this subtracts anything from the unacceptability of such an issue to begin with.

The control panel does allow that, but high FPS has never killed any card. Even if some cards have died, high FPS has nothing to do with it.
Please don't let this be another stupid PC myth that will never go away.

It's more that uncapped, simple GPU load with low load elsewhere on the card can expose flaws in power balancing

Hi to all the gurus out there :)

I see a VBIOS Update in the Precision X1 software.

Ya think I should go for it?

GeeBeeHi to all the gurus out there :)

I see a VBIOS Update in the Precision X1 software.

Ya think I should go for it?

Likely for rebar, so only if you want/need that. Pretty much unrelated to this.

MusselsIt's more that uncapped, simple GPU load with low load elsewhere on the card can expose flaws in power balancing

As you can see in the original Reddit thread, the people who claims their card have died have all throughout gameplay, not primarily in the menu. It's the high FPS which people think is abnormal which has led people to thinking it's dangerous, even though it's not.

High FPS in menus, loading screens etc. has been normal since the late 90s.
I mean no disrespect here, but if you knew how GPUs were programmed, you wouldn't even propose this idea. Rendering works by submitting a queue of commands to a driver, which in turn translates it into the GPU's native API. The GPU is in control over frame rate and potential frame pacing. A rendering engine submits new commands as soon as the driver accepts it, and is unaware whether the command is executed or if the GPU just accepts more commands. If you e.g. have V-sync enabled, the renderer will stall when it calls to swap buffers, until the driver detects the GPU is done. Attempting to time rendering from the rendering engine will be much more crude, as it has no direct control over when the frame is actually rendered. This is how games and thousands of programming examples works. Attempts to add sleeps or wait-loops in the rendering engine will be a very imprecise way to reduce frame rate, and is likely to cause stutter and input lag in a menu.

Whenever you see people claiming high FPS is dangerous, that a game needs to regulate power or balance the GPU, or even regulate the clock speed of the GPU, it's a clear sign they don't know what they're talking about.

Myth Busted!!

efikkanAs you can see in the original Reddit thread, the people who claims their card have died have all throughout gameplay, not primarily in the menu. It's the high FPS which people think is abnormal which has led people to thinking it's dangerous, even though it's not.

High FPS in menus, loading screens etc. has been normal since the late 90s.
I mean no disrespect here, but if you knew how GPUs were programmed, you wouldn't even propose this idea. Rendering works by submitting a queue of commands to a driver, which in turn translates it into the GPU's native API. The GPU is in control over frame rate and potential frame pacing. A rendering engine submits new commands as soon as the driver accepts it, and is unaware whether the command is executed or if the GPU just accepts more commands. If you e.g. have V-sync enabled, the renderer will stall when it calls to swap buffers, until the driver detects the GPU is done. Attempting to time rendering from the rendering engine will be much more crude, as it has no direct control over when the frame is actually rendered. This is how games and thousands of programming examples works. Attempts to add sleeps or wait-loops in the rendering engine will be a very imprecise way to reduce frame rate, and is likely to cause stutter and input lag in a menu.

Whenever you see people claiming high FPS is dangerous, that a game needs to regulate power or balance the GPU, or even regulate the clock speed of the GPU, it's a clear sign they don't know what they're talking about.

trust me, i had a GPU with insane coil whine for a few years when the load went unbalanced - if you get that in menus and load screens and it causes damage, it doesnt have to kill it in the load screens to have been DAMAGED during them

My melted out PCI-E power cables didnt happen in one quick burst, it was a slow build up of overheating and melting, and is the most likely cause of what happened here, too
Left on menus for ages: damage simply happened much faster

Again, this is all about exposing a flaw in the GPUs via slow death, not about "high FPS is insta kill"

Musselstrust me, i had a GPU with insane coil whine for a few years when the load went unbalanced - if you get that in menus and load screens and it causes damage, it doesnt have to kill it in the load screens to have been DAMAGED during them

Coil whine isn't dangerous by itself. It can sometimes be a symptom of bad power delivery, which in turn can lead to damage, but coil whine resulting from high FPS is harmless, even if you do it for hours at the time.

MusselsMy melted out PCI-E power cables didnt happen in one quick burst, it was a slow build up of overheating and melting, and is the most likely cause of what happened here, too
Left on menus for ages: damage simply happened much faster

Melted PCIe power cables means either the card does not follow the spec (defective design), you overclocked the card or the cables were not up to spec. Either way, high FPS or high load in games has nothing to do with it.

MusselsAgain, this is all about exposing a flaw in the GPUs via slow death, not about "high FPS is insta kill"

There is neither a statistical basis nor a technical basis for claiming why high FPS should damage anything, either slowly or quickly. High FPS has happened for years and it has never resulted in any damage. The "high FPS is dangerous" claim is beyond laughable at this point.

The cards that are failing are either defective and on the brink of failing, or they have been overclocked or mined on. It doesn't matter if you play this game or another comparatively demanding game, your card is just as likely to fail during load anyway. Hundreds of cards are failing every day during gaming, the only reason why this game is blamed is that people are more likely to suspect a recent change as the culprit, especially something in alpha/beta.

efikkanHundreds of cards are failing every day during gaming, the only reason why this game is blamed is that people are more likely to suspect a recent change as the culprit, especially something in alpha/beta.

Is there a daily tally somewhere that tracks this?

efikkanCoil whine isn't dangerous by itself. It can sometimes be a symptom of bad power delivery, which in turn can lead to damage, but coil whine resulting from high FPS is harmless, even if you do it for hours at the time.


Melted PCIe power cables means either the card does not follow the spec (defective design), you overclocked the card or the cables were not up to spec. Either way, high FPS or high load in games has nothing to do with it.


There is neither a statistical basis nor a technical basis for claiming why high FPS should damage anything, either slowly or quickly. High FPS has happened for years and it has never resulted in any damage. The "high FPS is dangerous" claim is beyond laughable at this point.

The cards that are failing are either defective and on the brink of failing, or they have been overclocked or mined on. It doesn't matter if you play this game or another comparatively demanding game, your card is just as likely to fail during load anyway. Hundreds of cards are failing every day during gaming, the only reason why this game is blamed is that people are more likely to suspect a recent change as the culprit, especially something in alpha/beta.

we've seen this before, furmark. you're missing the historical precedent for this exact thing.