Dead EVGA GTX480

[=[PULSAR]=]

Well I have have my GTX480 for about a month now, since the beginning of May pretty much 24/7 on gpugrid with fan at 100%. I was just on firefox checking the forums and the computer just shutdowns and my room fills with the smell of burnt hardware. Open my case and its even worse, remove the gpu kind of had an idea that would the culprit and it smelt like something definitely got fried. Replaced it with a GTX295 and no issues. It sucks I now have to RMA a card that only had a month of use on it and have to pay the darn shipping costs Not to mention the more than likely slow turn around time from evga.

[hlonipha]

that sucks

[Movieman]

980x..... dead
GTX480..dead
and you think I'm letting you near a SR2 board with $3000 in cpu's in it?

Sorry to hear pal..

[=[PULSAR]=]

980x..... dead
GTX480..dead
and you think I'm letting you near a SR2 board with $3000 in cpu's in it?

Sorry to hear pal..

I know I'm having crappy luck as of late, the 980x okay w/e I'm glad I got the money back for it more money for the sr-2 build, the GTX480...that thing wasn't even overvolted and excessively cooled, temps never exceeded 60C. I'm not gonna lie Dave I'm pretty scared to build my SR-2 now

[Movieman]

I know I'm having crappy luck as of late, the 980x okay w/e I'm glad I got the money back for it more money for the sr-2 build, the GTX480...that thing wasn't even overvolted and excessively cooled, temps never exceeded 60C. I'm not gonna lie Dave I'm pretty scared to build my SR-2 now

Nah, no need to be scared at all.
I have all the good "safe" settings and your chips are beasts.. I mean beasts and thats at stock speeds.
At 4000+ they will amaze you..

[Andrea deluxe]

more 480 dies....

probably vrm fail......

[=[PULSAR]=]

more 480 dies....

probably vrm fail......

I concur if I shove my face onto the card and try and pinpoint the smell its right above the VRM area. It smells more towards the intake fan than the exhaust. I would take the card apart but its going for RMA so why bother.

[WhiteFireDragon]

ah that really sucks. now there will be even less fermi cards on the grid. where did you buy the card? if you bought it under 30 days ago, you may be able to get a replacement from the store rather than going directly through EVGA. if it's newegg, they'll pay both way shipping. tigerdirect might, depending on which CS you talk to.

[=[PULSAR]=]

Its like 2 weeks over the 30 days. I didn't know newegg paid for both ways on shipping.

[sRHunt3r]

Wow Pulsar sorry to hear it man; here's to a quick turn around

I would also push the issue of advance replacement.

[evoted]

sorry to hear your losss

good luck with the RMA

[Razrback16]

Sorry about your luck on that. I had one of my 4890s blow a VRM chip a couple weeks back. At least yours will be an easy RMA process given it's so new.

[WhiteFireDragon]

i bought tons of stuff from them, only 3 things went bad and if you talk to a newegg rep online by chat, they will give you a return shipping label that's been prepaid, and also ship you the new item within a day after receiving the defective one. i guess it doesn't matter since it's 2 weeks over like you said . sorry to hear about this loss, it's not a small one either.

[saaya]

so what, they are not designed to be under 100% load?
sorry to hear man... if its just the pwm they should be able to fix the card, so maybe turnaround will be fast...
i thought evga was good when it came to warranty stuff though?

[=[PULSAR]=]

I have only had to RMA one evga product before and their hold time on the product was long roughly 2 weeks for a DOA board. I guess we will see I submitted and RMA request earlier today so should have a RMA in the next 24-48 hours. and saaya gpugrid doesn't even run 100% load it only runs it at 70% Really apprehensive about running a Fermi on LN2. I tried removing the stock heatsink because I was bored and couldn't the screws were so tight I was just stripping them, so I gave up.

[death]

dam its strange thow were u said u smelt it was it around here i no this is the 470 but was it about the same place as i have mine folding as we speak and dam that spot is like red hot ive added a small tall heat sink to drop the temps and u right that's about the place just under it that the vrms are and mine is on ek full block

[=[PULSAR]=]

Yeah thats exactly where the smell was coming from. I have no doubt that its the vrms.

[Wiggy McShades]

Yeah thats exactly where the smell was coming from. I have no doubt that its the vrms.

did you have a SC card by any change? my first SC lasted 4 days (knew it wouldn't last when 2mhz higher on memory clock caused it to artifact) and now i just got one back from rma and it looks to bed another one not long for this world (already started to artifact at stock). I honestly think these 480's are going to have a very high failure rate.

[p2501]

Is that so? We only know of two dead 480's here, both died from burned up FETs. I think Pulsar and the other guy just had bad luck.

Pulsar: hopefully your RMA goes well!

[=[PULSAR]=]

did you have a SC card by any change? my first SC lasted 4 days (knew it wouldn't last when 2mhz higher on memory clock caused it to artifact) and now i just got one back from rma and it looks to bed another one not long for this world (already started to artifact at stock). I honestly think these 480's are going to have a very high failure rate.

No it wasn't just a regular version, it was a pretty good clocker. Was running it at 800/1600/2000 24/7 had it up to 850/1700/2000 for folding when I first got it, then temps in the house starting going up.

UPDATE:RMA got approved and sent it in today. It was funny when I was on the phone with RMA support, they were like what seems to be the probelm. "Ummm my card started smoking and smelt like burning electronics."

[Wiggy McShades]

Is that so? We only know of two dead 480's here, both died from burned up FETs. I think Pulsar and the other guy just had bad luck.

Pulsar: hopefully your RMA goes well!

I've just seen a lot of people having trouble with the 480's recently, maybe I'm jumping to conclusions. You have to admit it wouldn't be too ridiculous to think a rather hot running card could run into problems.

@pulsar: that card was a beast, did you hold a funeral ?

[Onoff312]

dam its strange thow were u said u smelt it was it around here i no this is the 470 but was it about the same place as i have mine folding as we speak and dam that spot is like red hot ive added a small tall heat sink to drop the temps and u right that's about the place just under it that the vrms are and mine is on ek full block
]

I think its pretty obvious and a design flaw clearly for them to put hot components on an area that gets no airflow due to being under the fan motor.

[p2501]

I've just seen a lot of people having trouble with the 480's recently, maybe I'm jumping to conclusions. You have to admit it wouldn't be too ridiculous to think a rather hot running card could run into problems.

I wouldn't think of it as cause/effect scenario. You have a manufacturer that knows it's new baby will run bloody hot so they spec the components according to the needs - with a higher tolarance to heat. In which case the card would just do what it can do best, blast out heat, with no ill effect. That is, if we leave the laws of economics out of the game.

If we see defects, maybe it's just the normal propability of even a correctly spec'd component to die. We wont know for sure without statistical evidence.

[=[PULSAR]=]

Looks like evga RMA service was top notch this time around

They received it on 6/11 and shipped replacement on 6/14. Turn around was 2 business day not bad at all.

[570091D]

I wouldn't think of it as cause/effect scenario. You have a manufacturer that knows it's new baby will run bloody hot so they spec the components according to the needs - with a higher tolarance to heat. In which case the card would just do what it can do best, blast out heat, with no ill effect. That is, if we leave the laws of economics out of the game.

If we see defects, maybe it's just the normal propability of even a correctly spec'd component to die. We wont know for sure without statistical evidence.

i have to wonder if nvidia didn't build the voltage regulation circuitry with enough reserve for the chip, evidently even they didn't know the finished product would pull 250 watts. this strikes me as odd, as usually nvidia boards are well engineered, it's the silicon that fails...