FloodControl

kevinc313

dvdood

kevinc313

dvdood
These might be solid advices what to do to monitor/preserve card, however anyone buying gpu shouldn't be even concerned about it breaking is such a short period.

Stick to cards under 200w.  GTX 1660 for you.

Now ur being mean, like i said anything wrong.

My man.  This is not like a PS5.  The 3090 FTW3 has a STOCK, OUT OF THE BOX POWER LIMIT of FOUR HUNDRED AND TWENTY WATTS.
 
Anyone who buys one (for ~$1800) should be *EXTREMELY CONCERNED* about setting it up properly and not breaking it.

All that being said, as Sajin has pointed out there is ample evidence of factory issues with the 3090, at least in its FTW3 variants. It definitely falls on EVGA to design a full card around the chip that functions to reasonable expectations: PCI spec, adequate cooling, and general quality of build. So far, we've seen plenty of reports of the power delivery not complying with PCI spec (which itself is an issue that can damage hardware) as well as inadequate cooling of VRAM on the backside of the card. The obvious remedy to these problems is for EVGA to revise the power delivery of the card to load balance properly to PCI spec, and to give the VRAM adequate cooling with a more capable backplate. They've already done step one. EVGA has been at fault before (1070/1080 thermal pad issues being the biggest most recently) and there's no doubt that they could have missed something here, too. They dodged the bullet on the initial capacitor issues that other AIBs ran into on Ampere launch, but they aren't perfect nor do they need to be defended as if they were. You aren't their legal counsel.
 
Off-handedly ruling out manufacturer fault and blaming customers for not going to the extreme beyond the stated requirements of the card is not how to address the overall problem. Nothing VegasBaby has stated points to any incorrect installation, and they shouldn't be treated vindictively for wanting more info on what could be causing their issues. All we can do is work with the info given, and leave the rest between them and EVGA's RMA process.
 

Oh, IMHO the 3090 FTW3 should not exist.  The power limit out of the box for an air cooled card with VRAM on the back is FAR too high.

hashmeir70
Lmao that Kevin dude is a trip.

Gaming PCs are not hard.

Just because Kevin wants a NASA control panel to operate and monitor a gpu doesn't mean you need it.

Suggesting you can't drop in a 3090 and just run it. Is blatant misinformation. Hell, even windows will fetch the drivers for you. As long as you have the recommend power supply you're fine. Even poor cooling is okay, it will just down clock and be the world's slowest 3090. By design you cannot and at the very least should not be able kill a gpu with the settings NVIDIA (not evga) allows the user access too.

OK buddy, no more complicated than a toaster oven or vacuum cleaner I suppose.

typical and usual fanboys... 

farmer_ta
After my FTW3 died 5 months ago I still monitor the forums for any updates regarding a fix or change to the design of the FTW3 series, it seems the issue is still here sadly. I am convinced these cards have the same reliability as a Chrysler Sebring. Blaming the OP saying his/her setup is "inadequate" is pathetic. These are consumer cards and many have said already, they are plug in and play. It is just poor engineering from EVGA if bad airflow or "lack of monitoring" can cause GPU failure. 
If I had an expensive Mercedes that explicitly takes 91 octane gas and put 87 octane gas in it, will it get damaged ? The answer is no, cause it was actually well designed and the cars ignition system can compensate for it. It doesn't even matter if OP runs his card with a no name brand PSU, the card should be designed that even poor power delivery shouldn't damage it ! 
Good luck OP I hope your issue gets solved. I have seen someone on this forum get a refund after constantly receiving bad cards so maybe you can try that. 

Thank you. I spoke with EVGA, have an RMA going. They're paying for shipping costs. 
 
These trolls had me thinking I was crazy for a second there...

kevinc313

hashmeir70
Lmao that Kevin dude is a trip.

Gaming PCs are not hard.

Just because Kevin wants a NASA control panel to operate and monitor a gpu doesn't mean you need it.

Suggesting you can't drop in a 3090 and just run it. Is blatant misinformation. Hell, even windows will fetch the drivers for you. As long as you have the recommend power supply you're fine. Even poor cooling is okay, it will just down clock and be the world's slowest 3090. By design you cannot and at the very least should not be able kill a gpu with the settings NVIDIA (not evga) allows the user access too.

OK buddy, no more complicated than a toaster oven or vacuum cleaner I suppose.

Honestly not really. A user can absolutely plug it in and go. Can you add more steps if you want to delve deeper? Of course. How hard is it to insert something that connects like a giant stock of ram with 1 screw and 3 power cables where everything goes only one way. after that Boot, windows will fetch the drivers and NVCP and then you play. The bare minimum is plenty.

Calm down Rube Goldberg

Kevin why are you so adamant on user error? Can't cards have design flaws, break?

Got mine in October, haven't had any problems and I game everyday. I've actually been really lucky, been using Evga cards for a long time and never once had a failure, knock on wood.
 

Who is this Kevin Guy? Why so toxic? Who hurt you my dude?

Stumbled upon this thread by chance and boy what's with the $hitty responses given to the OP?
On the subject of "you killed your card". No. None of the users should be expected to tinker with their cards whether through software or hardware to keep the card from dying. There should be safe guards built into the card for that. And to my understand there is. So unless the guy shunt resistor mod his card to death it's not his fault. Period.

Okay, most of what I'm stating here is MY OPINION.  You can like it, not like it, whatever.

First: 3090's have some cooling issues as most of them are mounting half the RAM on the back of the board, which can lead to thermal issues on the PCB itself when both sides of the card are being heated by RAM.

Honestly, if I picked up a 3090, I'd either detune the thing down to mitigate power draw and thermals, or I'd watercool the whole damn thing, front and back.  PERIOD.

Second: The 3090 is an ENTHUSIAST card.  Not a LUXURY card.
If we were talking automobiles, a 3070 is a V6 Dodge Challenger.
A 3080 is a Hellcat.
A 3090 is a Demon/Redeye
A 3090 is NOT a Mercedes S-Class.

You are getting progressively MORE specialized hardware, running closer to the edge of the device's operational envelope.
One doesn't buy a Demon and then take it out rock-hopping.  Well, unless you're willing to trash the very rare, EXPENSIVE vehicle.

As such, you need to PAY ATTENTION to what you're doing.  And for a $1800 investment in a single video card, you SHOULD be monitoring it.

If you want "set it and forget it", get a 3070 or 3080.

Now that's NOT saying it's VegasBaby's fault.
Quite simply, we don't have enough information to make that assessment.
But the more we monitor, the better of an idea we have of what's going on when the device fails.
And it means that better feedback can be provided to EVGA from the get-go.  Speeding RMAs, and giving them good technical data they can use to troubleshoot their product.

z1nonly
 

firerain
 
No, The product needs to be made that will last with its specs. Psu, airflow, framerate should be irrelevant if in spec.

 
This.
 
Put the requirements on the box and engineer the card to work within those specs. If the card requires a PSU forged from Unicorn hooves, put it in the requirements. Don't tell your customers to use x-wattage PSU with y-efficiency rating and then, after the card dies, say the PSU is not the "chosen one".
 
Most manufacturers understand, and engineer for, the various scenarios end-users will put their products into. 
 
A 3090 with "special needs" discovered by community trial and error, and itemized on discussion forums, is a 3090 with issues.
 

I agree with this mindset too...

Cryptospook

VegasBaby

kevinc313
- Quality PSU
 
- Good Air Flow
 
- Monitoring

... I was playing Halo ...

I've seen like more than a dozen people complain about their 3090 giving up the ghost while playing Halo MCC. Something way wrong with that game...

Something way wrong with the card that it can't handle a game.

KingEngineRevUp

Red_Infern013
Stumbled upon this thread by chance and boy what's with the $hitty responses given to the OP?
On the subject of "you killed your card". No. None of the users should be expected to tinker with their cards whether through software or hardware to keep the card from dying. There should be safe guards built into the card for that. And to my understand there is. So unless the guy shunt resistor mod his card to death it's not his fault. Period.

It's a 5 week old dead thread. What's the point of reviving it and calling out users who have already been banned? 

A hundred post count?

You REALLY needed to ask? 

KingEngineRevUp

Red_Infern013

It's a 5 week old dead thread. What's the point of reviving it and calling out users who have already been banned? 

Came across it while browsing and the responses rubbed me the wrong way, especially the attitude of "you have to be a tech wizard to deserve the tech you buy" that's been going around lately, not only from users but from manufacturers as well who ship products in a not-so-adequate out of the box settings. If you buy a laptop nowadays most likely it's gonna overheat and throttle with out of the box settings. Discrete GPUs with not so great cooling as well will do so and CPUs. It's not like it's a year old post and I thought I'd leave my opinion here instead of starting a new thread, to my understanding that's always preferred.
 

charlesborner

KingEngineRevUp

Red_Infern013

A hundred post count?

You REALLY needed to ask? 

Did you just assume why I'm posting? *TRIGGGGEERRRD* but nope, don't need the 100 posts and I'm way above the 100 posts if I'm not mistaken. I only joined recently but I actually enjoy being part of the community here. I don't like reddit and I don't frequent other tech forums.
 

Yea my first card just flat lined 2 weeks ago (red led's of death).  I have great airflow, 1600W EVGA Supernova PSU, and my 2080ti worked totally fine without issues since I had it.  I am on my second card now and "so far" so good.  Here's hoping...