GPU Fried, was it my GPU or PSU?

Ghostmb

Reputable
May 31, 2015
4
0
4,510
Hey guys, first time poster here so I'm not really 100% on the correct way to structure this post, but here goes.

Components in question -
GPU - MSI Twin Frozr GTX 780 (19 months old)
PSU - Corsair RM850 (5-6 months old)

I was playing Battlefield 4 the other night, all of a sudden my PC shuts down, no artifacting, errors or other warning, and then I smell a very distinct burning smell from the case which fills the whole house with an unsettling aroma.

Needless to say, the PC wouldn't boot. The next day I started trouble-shooting the problem, suspecting that the PSU had gone. I removed the GPU, sound card, and one stick of RAM. To my elation the PC booted no problem, after putting in the RAM and sound card it booted fine too. This obviously led me to believe that the GPU was indeed the component that got toasted. A quick whiff through one of the fans on the heat sink and I could smell that distinct burnt out smell.

The question is, why would this happen? Don't GPU's have some built-in shut-down temperature to avoid damage if the cooling fails? I also find it weird that there was no artifacting or other tell-tale signs as the GPU getting hotter. In any case this is my second Twin Frozr card, I still have the older 580 running in a secondary machine and going strong, and have had no problems with the 780 up until now.

Temperatures on the 780 have never gone above 80C, and my 4790k sits at 60C under load, neither are over-clocked past the settings they came with.

I'm going to RMA the card tomorrow through my retailer, since it has a 2-year warranty. However, I was thinking of unplugging my PSU unit, along with the wire that connected to my GPU and letting them take a look at it, I don't have a multi-meter. I can't say I've had any problems with the PSU, and I've been loving it so far, but is it possible that this was the PSU's fault? The last thing I want is to get a replacement card and have that fried some-time during the next few weeks or months by the PSU.

Any thoughts/expert advice would be appreciated, I've been building my own PCs for yours but still consider myself an amateur. Cheers!
 


Yeah, I suppose so. Still, I don't know what would make it go from a 120fps gaming experience to insta-melted.

 


So what you say I have no need to worry about the PSU? Since everything seems to be working as normal with integrated graphics.

 



Well, in these cases there is ALWAYS that rare chance that the psu somehow gave a sudden surge to the card from the rail, frying the component.

It's hard to confirm if that happened. So you can only RMA that gpu and hope it won't happen to new one.
 


Right ok, how can the same shit happen to the same guy twice? Right? 😉

Just for future reference, I've heard it's safer to use 2 cables to connect a single GPU, one for the 8-pin and one for the 6-pin, rather than the same cable with a splitter, any truth to that?
 


Yes and no.
It will still give same amount, but it will put stress on that cable.
Possibly causing power surges and cable burning.

So basically it is safer for whole HOUSE to use 2 separate ones.

Oh and of course, some gpus require 2 DIFFERENT size cables. 6 pin + 8 pin. So splitter wouldn't give right amount of pins. Thus leaving gpu under-powered.
 
I CREATED AN ACCOUNT JUST TO COMMENT HERE.

The same thing happened to me just recently. May 29th 2015.

GPU: EVGA GTX 780
PSU: Corsair CX750 (Both are 8-9 months old)

I was playing Tower of Guns, not even that demanding, just waiting a friend to play cs:go later.
Around 6 pm, suddenly my pc shuts down, like the power was cut, no warnings, no signs.

After i spent the weekend diagnosing the problem, i found out that the video card was burnt, electrical burn smell came from the power connector area.

There is really a rare random chance of simply shorting out for no apparent reason? Then i had real bad luck.

I feel like a family member just died from a heart attack, i had an emotional connection to that GPU. 🙁