liud21
first it was "YoUr PsU iS nOt GoOd eNufF" now its "YoU cAnT pUsH a 3090 tO iTs LiMiTs, lIkE tHaT". If a Top of the line GPU cant play a game with default settings, or High settings, there is something wrong with that GPU.... The only thing to blame now is EVGA, they for  sure cheap out on a certain part within their 3090 FTW3.... LMK when ASUS or even ZOTAC has a Reddit thread with over 1k comments about their 3090 dying from playing a Game with a GTX 970 recommended requirement....

deathlessdd

kevinc313
Here's the first anecdotal account of out of the normal operating parameters I've seen, vram junction temps of 115C:
 
 
 

the 3090s have vram on the bottom that runs at 100C during normal gaming, and someone did report seeing 115C during new worlds on his ftw3, a majority of the bricks seem to be evga but there are reports of abnormal temps and power surges even on other hardware. Though the bricks seem to be caused by something else (signs points toward extremely high fps in the menus, causing a power surge in an already flawed card model), 115C vram temp doesn't help. My old gpu had massive coilwhine over 200fps, with the increased horsepower of the 3090, going over 9000 fps in the menus might be a condition that can pop a cheap capacitor consistently.

 Of course, I would like to see any other accounts of how the card was operating at the time of failure. 

Could the vram temp be directly connected to the to killing the gpu though, Im not even sure if high frame rate cause even cause high temp on vram though.

kevinc313

deathlessdd

kevinc313
Here's the first anecdotal account of out of the normal operating parameters I've seen, vram junction temps of 115C:
 
 
 

the 3090s have vram on the bottom that runs at 100C during normal gaming, and someone did report seeing 115C during new worlds on his ftw3, a majority of the bricks seem to be evga but there are reports of abnormal temps and power surges even on other hardware. Though the bricks seem to be caused by something else (signs points toward extremely high fps in the menus, causing a power surge in an already flawed card model), 115C vram temp doesn't help. My old gpu had massive coilwhine over 200fps, with the increased horsepower of the 3090, going over 9000 fps in the menus might be a condition that can pop a cheap capacitor consistently.

 Of course, I would like to see any other accounts of how the card was operating at the time of failure. 

Could the vram temp be directly connected to the to killing the gpu though, Im not even sure if high frame rate cause even cause high temp on vram though.

Well, the vram does take power. It uses more power depending on what kind of load it is under and what kind of overclock it has.  Like most silicon it takes more power to do a task the hotter it gets.  24GB is alotta vram.  For it to get to 110C+ quickly it has to be drawing a ton of power under a high load.  Is a high framerate, chucking a bunch of data in and out of vram, something that could cause high power use?  Seems like a reasonable theory.

liud21
so like 90% of the cards are EVGAs, the rest are a mix of AMD/Nvidia AIBs... lol. EVGA, just come out and own up.... They probably email Jay, Bro you gotta chill, or youre not getting the SUPER cards when it comes out...

ty_ger07

kevinc313

deathlessdd

kevinc313
Here's the first anecdotal account of out of the normal operating parameters I've seen, vram junction temps of 115C:
 
 
 

the 3090s have vram on the bottom that runs at 100C during normal gaming, and someone did report seeing 115C during new worlds on his ftw3, a majority of the bricks seem to be evga but there are reports of abnormal temps and power surges even on other hardware. Though the bricks seem to be caused by something else (signs points toward extremely high fps in the menus, causing a power surge in an already flawed card model), 115C vram temp doesn't help. My old gpu had massive coilwhine over 200fps, with the increased horsepower of the 3090, going over 9000 fps in the menus might be a condition that can pop a cheap capacitor consistently.

 Of course, I would like to see any other accounts of how the card was operating at the time of failure. 

Could the vram temp be directly connected to the to killing the gpu though, Im not even sure if high frame rate cause even cause high temp on vram though.

Well, the vram does take power. It uses more power depending on what kind of load it is under and what kind of overclock it has.  Like most silicon it takes more power to do a task the hotter it gets.  24GB is alotta vram.  For it to get to 110C+ quickly it has to be drawing a ton of power under a high load.  Is a high framerate, chucking a bunch of data in and out of vram, something that could cause high power use?  Seems like a reasonable theory.

And that is the owner's fault?

deathlessdd
Class action Lawsuit incoming? Game kills 100s of 3090s for unknown reasons...

Sajin

_Gir_
Did EVGA ever make a comment about the red light issue?  I though Jacob implied he would some months ago.

Haven’t seen anything.

MaxTheOwl
Hello,
 
I will add some of my experiences.
 
My setup: Intel Core i9 10900K@5.1GHz, Gigabyte Z490 Aorus Master, G.Skill Trident Z 3600MHz, be quiet! Dark Power Pro 12 1500W, EVGA RTX 3090 FTW3 Ultra (PCB Rev. 1.0/BIOS 94.02.42.80.27) PL@119%/Core+100/Memory+200/Voltage+100, Framerate capped to 160FPS via PX1 all the time @1440p - These are my daily gaming settings
 
I played "New World" yesterday and today for a total of 8 hours and everything was normal for me. Some say it is the high utilization with the high fps that causes the death of the cards. But I don't know. I mean, I play Battlefield 4 with a average of 499 Watts and 100% GPU utilization without problems for hours. I only see maximum wattage in the menu of "New World". Ingame, "New World" could rarely use all the power of the card, so it is strange why this guys https://www.youtube.com/watch?v=n6sYOqPPTN0 GPU died ingame (with low fps like it seems).
 
It would be good to know if only the cards with the old PCB Rev. 0.1 have this problem.
 
Kind regards.
 
EDIT: Memory Junction never goes above 90s in "New World"
 
 

xtremegamer
Based on that statement The game is killing the cards unbelievable..

Eisai_Kurosawa
Tried to run this game for 20 minutes, I have maxed out graphics settings on the main menu, turn off V-Sync, create new characters, then play around for about 10 minutes.
I tried both stock clock settings and my undervoting overclock settings, Luckily my 3090 is still alive.
 
My spec:
5900X
3090 FTW3 Hybrid Ultra(stock settings with no FPS limit)
Revision: IDK
Driver: 471.41
BIOS: 94.02.42.C0.11
4K@144hz Display with HDR enabled
32G 3600mhz RAM
1000W Seasonic 80+Titanium PSU
OS: Windows 11 22000.71
 
Seems this game have some scenario cause GPU utilization from 0% to 100% in a second, also GPU is running 76C in New World, it's couple degrees hotter than other games. I think this is still totally fine for a game. I don't think a game was able to kill my graphics card, Only graphics cards can have design flaws.