Discussion in 'Frontpage news' started by Hilbert Hagedoorn, Jul 22, 2021 at 8:26 AM.
Thank you for your review,
you just earned a free month of Amazon Prime!
No, it really is a hardware issue. This isn't the first or even 2nd time Nvidia has had parts die prematurely because of a stressful workload. The chip really ought to be able to protect itself from such things. That's what thermal throttling is for.
I have an EVGA 1080 SC not a single issue, not even with this game
Wonder if it's the vram thats getting too hot.
Of course they have. And it is affecting more titles than our frtc. I think all people should limit their frames up to their monitor HZ or ~25% more as a max value.
It's not "killing 3090's" It's killing EVGA's FTW3 cards that have been known for months to have power delivery issues, and have been dying left and right long before this game came into the picture.
Sold my FTW3 after Evga kept ignoring the PCIE power issue, dodged a bullet
The problem is not just with the 3090s, and I speak from first-hand experience.
On my main PC I have a 6800XT (ASUS TUF) and I noticed immediately while queuing to access the server that the card was getting warmer than normal. I closed the game, then set the FPS cap to 60. Now the temperature in game is 65°C max.
On my office PC I did the same this morning, mostly as a precaution for what I'd heard about the 3090, but it didn't help.
The result is that the ASUS R9 290, which has always worked fine now lies on the desk and still smells bad.
It seems is an evga rtx 3090 design flaw :/ .
Might not be exclusive to Evga but it's shining a light on their flaws, and Amazons!
EVGA has probably sold more 3090's than anyone else too. Would make sense they had a bigger representation of failures.
Interesting to see AMD cards in the mix now too.
The issue is with poorly designed graphics cards. And it seems to be exclusive to the top end. Which I don't understand? Mid range cards are better designed than top end?
Also, there's a difference between "shut downs" and failures, Jay!
Such issues should be prevent by the firmware, not by users or applications.
JayTwoCents never wrote a line of code in his life I am ensured now. He knows everything.
The same goes with people blaming developers and Amazon. The fault is at hardware or better placed firmware level. Vendors and AIBs to blame. Safety precautions anyone?
Glad i decided to always lock my fps to the max refreshrate of my displays.
A huge one, specially when the card shuts down for protection(like it should)
Warhammer 40.000 Dawn of war 3, has the same menu problem.
I have used it for GPU stresstest a couple of times.
My Vega runs 120 FPS V-sync locked, 220W chip power and 1550Mhz full boost speed in the main menu.
Apex Legends also has a fairly high load in a rather simple main menu.
I've reported on this ages ago as i had a problem with a game where it would shoot up to between 3000 and 4000 fps while showing cut scenes or in menus and it caused my system to go nuts. Lucky for me it never damaged my gfx card. BTW that's with V-Sync on and frame rate locked to 240hz/fps. It still did it but soon as the scene switched or the video stopped it would go back to normal. It was a while back now and i forgot which game it was.
I think it was in Doom Eternal and while in the menus too. My pc just went mental and went from 35db up to like 80db and it frightened me at first when it happened. I think it caused my gfx card to throttle because i heard it step down a notch and then went back to normal once i went out of the menus.
I also think a game can't break a gfx card but it can show up weakness in a cards ability to cope with such massive increase in fps. This is why those whom experience this can get a warranty on their card. At least with 3090's you should have at least two years of warranty left or three if you signed up for it.