EVGA 2080 Ti XC burst into flame !!

I'm just going to take this as another sign that they know AMD is finally nipping at their heels, that they would put out such an obvious Pre-overclocked not consumer friendly product. Between ram failures, bios failures, driver issues, and now hidden flamethrower capabilities I'm literally just here with the popcorn meme waiting for their heads to implode.

Here's hoping the 9980xe and other 9th gen Intel chips don't suffer the same calamitous failure rate.
 
rebel2302 said:
i know EVGA has support 24 hrs and rma is really good too.
Click to expand...

More like used to, after the Cryptomining craze, they even removed their tool so second hand buyers can check about the warranty on their cards. Now you need to send them an e-mail with serial numbers and shit so they can advise you if you have any warranty left.
 
Damn, that sucks, only thing I had flame out on me was my refurb Sansa Clip MP3 player looped around my neck cycling home from work.
 
That's just a sign you need to buy more, to save more!

1420430567747.jpg
 
I was only able to obtain the Zotac 2080ti, and was kinda bummed. She isn't nearly as pretty as my EVGA 1080ti, but it seems the crazy-hot scale applies to video cards these days. EVGA should hook you up after that fiasco.
 
CAD4466HK said:
The Note 4 was a great phone, I still use one. I think you probably meant the Note 7 witch was recalled due to it potently being a firebomb waiting to happen. But I agree that it was a good thing for the OP that he was at home and on the machine when it happened.

If I was an owner of a 2080Ti, I would NOT leave it running while not in use, unless you have great wildfire insurance like all those celebs in Cali.
Click to expand...
Yeah I had a Note 4. The gist of my thought first thing in the morning was convayed.
 
Looks like it was a short circuit, especially if the pc powered itself off, vs just the display going blank but the pc still running. As long as something from inside the case didn't touch it, such as a stray molex connector, it was probably a bad component.

EVGA Rma will take care of you.
 
trentchau said:
"Don't you guys have fire extiguishers?" - Blizzard Rep
Click to expand...

It took me a few moments before I got that. That was funny as hell.

OP you should sue both EVGA and Nvidia for pain and suffering. Tell them your hand was burned. This is completely unacceptable for a $1200 item, WTF, wow just wow!
 
I'll pull my card and take some close up pics to see if we can figure out if some "glue" exists there which could melt. Also would help to know what exactly is in that spot which could potentially ignite.
 
pendragon1 said:
so op, whats evga have to say?
Click to expand...

Just contacted EVGA and got a cross shipping RMA in process. Hopefully I can have another replacement by weekend to start off my new build.
It was suppose to be part of my new 9900k system, glad I didn't stick this one in there yet...... Amazon preorder with 1 month delay saved it.. ;)


Slade said:
I'll pull my card and take some close up pics to see if we can figure out if some "glue" exists there which could melt. Also would help to know what exactly is in that spot which could potentially ignite.
Click to expand...

I have no idea what was there, since its all melted and I couldn't tell. This PCB should be a reference straight from nVidia though.
 
damn shame that has happened to you:/


im curious would this have happened anyway if it had a waterblock on?
 
I don't think the heat alone fried it. I measured the idle temps with my infrared thermometer and found 112F was the highest reading I could get out of that area which while warm, was not anywhere near temps which could self ignite. The heatsink on both sides appear to be well clamped on my card allowing good heat transfer away from the board. I'm thinking defect in the workmanship here is most likely culprit or bad circuit design.
 
Probably a bad solder joint or a bead of solder where it shouldn't have been. Great manufacturing process by EVGA as usual.
 
Slade said:
I don't think the heat alone fried it. I measured the idle temps with my infrared thermometer and found 112F was the highest reading I could get out of that area which while warm, was not anywhere near temps which could self ignite. The heatsink on both sides appear to be well clamped on my card allowing good heat transfer away from the board. I'm thinking defect in the workmanship here is most likely culprit or bad circuit design.
Click to expand...

Bad capacitors? Don't those usually burst into flames?
 
There are no capacitors in that spot from what I can tell. The large chips are ferrite inductors and seems to be small solder points indicating resistors only in that area.

For an inductor to "spark" and flame, there would have to be some runaway current inducing a strong magnetic field. This would explain the system shutdown as there would have been a major drop across a voltage rail, possible 5V or 3.3V which would be enough to cause a fault to just shut off. The inductors don't seem to be the source of the heat: it looks like ignition happened around the resistors in the area. Possible bad solder joint or a short.
 
Last edited:
Amazingly enough, this fits with the bad capacitors thing Nvidia has put out.


If the current in the inductor goes too high, like if the capacitors aren't storing charge, or the current stays too high, the inductors saturate, and aren't inductors anymore, they're a piece of wire.

Sticking a magnet to one can do this, so keep the supermagnets away from your power supplies. :)


My bet is, something else shorted, and this melted/caught on fire as it tried to dump current into it; that's about what 300W does to the traces on a PCB.

The nice thing is the the FR in FR4 is "Fire Retardant", which means the epoxy in the fiberglass the pcb is made with is not burnable, IF YOU REMOVE THE SOURCE OF THE FLAME.

In this case, it should go out when the Power supply went down.

There's a Spec, 94V-O, and it says it has to go out within 10 seconds.

https://en.wikipedia.org/wiki/UL_94

I've done this test, it's fun, but we used a blowtorch as a flame source. :)

Sorry for your luck, dude, but this will "Light a Fire under their Asses" to get this fixed.

I'd bet we'll see a recall soon; they Do NOT want to pay for house fires, pain and suffering, and all that stuff.


A switching power supply is like a charge bucket; you fill a capacitor from a faucet, thru a spigot, but the spigot only goes on/off.

Each time you fill the bucket, the controller looks at the bucket, and when the level falls, it adds more from the spigot.

If the level is Way low, it turns the spigot on longer or harder, and if the bucket is empty, it overloads and blows the spigot off the pipe... :)

This is the blown up spigot, pipe, and bucket, lol.
 
I've had GPU's go up in smoke before to melt the AGP/PCI-E slots. This is not a new occurrence and is rather rare to happen, but it can happen. 2 cards out of 10K+ units sold over the last few months is still a relatively small sample size. I think EVGA had significantly more 1080GTX failures as a single AIB partner and potential failures out of all 1080GTX cards enough to warrant a whole cooling design change. I play with nvidia's enterprise and consumer level cards every day and know full well how much heat they can generate under a load. A bad part here and there and with the wattage they are pushing, there goes a melted/burnt out trace.

Look at your motherboard as a whole and think about the latest AMD/Intel parts also pushing 200-300W power under load and think of the cooling and VRM needed to power all that. Now look at your GPU pushing similar power through on a board 1/3 the size. It's more clear now that the part selection across the board will be much closer to tolerance levels than the motherboard part selection.
 
Last edited:
nVidia has overdone themselves and knocked it out of the park with this one, its become so hot its literally bursting into flames...

glad in the end everyone is OK... hopefully warranty covers this, if evga says no I'd call nVidia directly.
 
Holy shit. I would not be happy with that.
EVGA better have a fucking grovelling response to this when its reported.
This could have killed you all if it occurred while the PC was running and not being monitored

KILLED YOU ALLLLLLLLLLLLLLLLLLLL

Dramatic enough?
 
Every single computer component you use is technically at risk of fire. As in some previous posts, they are rated to a degree that under most situations once the psu cuts off and there is no power, the flames should go out rather quickly as the components are fire retardant.
 
You must log in or register to reply here.
Back
Top