Shroud's PSU explodes on live stream while playing New World 
.
https://gamerant.com/shrouds-power-supply-explodes-new-world-twitch-stream/
. https://gamerant.com/shrouds-power-supply-explodes-new-world-twitch-stream/
-
Competitor rules
Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.
New World still killing GPUs? Jaytwocentz vs Amazon part2
- Thread starter Grim5
- Start date
Remember that after Fermi Nvidia has had driver detect Furmark and automatically throttle/limit clocks of the GPU to protect card.
So any game which really pushes the GPU can push card to the edge.
not a particularly good sub-system (gpu) design then. Forceable misuse should be designed for.
Shroud's PSU explodes on live stream while playing New World
https://gamerant.com/shrouds-power-supply-explodes-new-world-twitch-stream/
UPS not PSU.
The article is rubbish... they even called the UPS a Universal Power Supply, where it should be an uninterruptible power supply. Guessing he used cheap batteries in it that exploded in the end and nothing to do with the game again. Just people creating clickbait all the time now.
Jayzywocents made sure to tell us in his first video that this wasn't just nbidia gpus as he had aload of amd gpu users get I touch with him. Just makes him look worse...
the only thing annoying about him is his consent videos about sticking overheating high end components into toasters
It's why I despise journalists. I dread to think where journalism will be in another 5 years when you see how bad it is today.
Buildzoid rambling about the cause of the dead 3090's.
He says that the general power delivery design from Nvidia is a bit poor and a *********** for the RTX3090. One of the big issues is that all RTX3090 cards have very high OCP, they are basically pre shunt modded from the factory to pull LN2 power draw and that's to accomodate the fact that Nvidia knew the cards to hit crazy high wattage peaks. the RTX3090 is the definition of "just wing it"
RTX3090's can have 1ms power peaks that is usually reserved for LN2 and hardcore Kingpin type cards with way overbuilt PCBs. Nvidia and AIBs have budgets to hit and are not experienced enough with cards that can require such power draw. Nvidias recommended power delivery design is a bit wonky and OCP is extremely high, combine with AIB inexperience and tight budgets = consumers have GPUs that at any time can run into a highly specific scenario where their particular GPU can no longer cope and blows.
He says that the general power delivery design from Nvidia is a bit poor and a *********** for the RTX3090. One of the big issues is that all RTX3090 cards have very high OCP, they are basically pre shunt modded from the factory to pull LN2 power draw and that's to accomodate the fact that Nvidia knew the cards to hit crazy high wattage peaks. the RTX3090 is the definition of "just wing it"
RTX3090's can have 1ms power peaks that is usually reserved for LN2 and hardcore Kingpin type cards with way overbuilt PCBs. Nvidia and AIBs have budgets to hit and are not experienced enough with cards that can require such power draw. Nvidias recommended power delivery design is a bit wonky and OCP is extremely high, combine with AIB inexperience and tight budgets = consumers have GPUs that at any time can run into a highly specific scenario where their particular GPU can no longer cope and blows.
Last edited:
Short version is cards using these controllers for the VRM/BUCK UPI 9511R and UPI 9512R made by uPI Semiconductor corp are the ones that have made the VRM go up in smoke so far and no other cards with different controllers have had this issue so far, example On Semiconductor NCP81611 and NCP81610 VRM/BUCK controllers used on 3090s on some manufacturers cards have had no failures mentioned so far, same goes for the Monolith MP2888A and MP2886 chips on the 3090 FE cards and the Asus 3090 Strix plus others.
Basically check a teardown of your card and hope you don't have uPI chips for the VRM controllers, but doesn't mean your card is bad, but so far only cards that have gone pop are cards with these chips, EVGA and Gigabyte use them and other AIBS, so not sure check a teardown.
He's basically saying he suspects the uPI Semiconductor corp VRM/BUCK controllers are the issue as did Igors Lab a while back too, but Igor thought EVGA disabled some of the protections on these chips and used the iCX stuff they have in their FTW cards and that may have caused some of the issues too.
He waffles a lot in the video but that is basically what he's trying to say in way too many words.
I have to also ask.. what if it's not just the uPI VRM controller and just a bad combination of controller with VRM chips that may also be causing this issue... Anyways no way to really find out unless you have a decent supply of cards to test them till they give up the magic smoke and then fix them with higher rated components on the VRM and again test and see if they pop again with the same controller and another test would be to keep the same VRM chips and use a different controller on the cards that went pop and see if they go pop again. This needs real testing but guessing EVGA know the answer as they swapped the FTW3 3080ti cards controllers to "On Semiconductor" ones ..from the 3080 and 3090s uPI ones... maybe more than just bad solder...
and the new 3090 FTW3 cards have the new controllers too and that's the real fix ?
Last edited:
Put voltage back to stock and run furmark for 24 hours.
Soldato
Kinda don't want to try this on my 3080ti, the GF is paranoid enough about the tower behind the TV, if she finds out the game blows components up she'll end up leaving.
The Zotac AMP HOLO 3080Ti you have is controlled by a MPS2886A and a MPS2888A controller from Monolithic Power Systems. So far no reports of these cards with these controllers going pop, FE uses them and Asus strix cards.
https://www.techpowerup.com/review/zotac-geforce-rtx-3080-ti-amp-holo-black/3.html
Last edited:
Buildzoid says there could be a lot of dead RTX3090 cards in another couple years.
Even the cards that don't have an issue with new world right now, because they are being run right at the limit of the power delivery components with these constant 500/600/700w transient spikes and these components don't last long like that.
In his experience with many cards, when they are run at max spec of the power delivery and power stages with these kinds of loads, you get about 3 years use before the card goes boom
Even the cards that don't have an issue with new world right now, because they are being run right at the limit of the power delivery components with these constant 500/600/700w transient spikes and these components don't last long like that.
In his experience with many cards, when they are run at max spec of the power delivery and power stages with these kinds of loads, you get about 3 years use before the card goes boom
Last edited:
