Alienware M17x R3 Brand New 7970m Died!

Hi all
And the problems with my laptop continue, here we go:
My laptop is a M17x R3 that used to have a 6990m and 8GB of ram, A12 bios
The 6990m decided it wouldn't accept drivers anymore. After me spending 2 days on it and a computer shop spending one day on it I gave up and bought a 7970m from ebay for $400 NZD and imported it from china.

Installed the card along with a extra 8GB of RAM while I was at it. Forgot to plug fan in for card, it worked flawlessly for 30min and then died. New GPU "not detected" in BIOS. Bought another 7970m from china binned the first one and installed it properly this time. (over $900 spent at this point eeek)

Eveything worked this time, installed card, disabled iGPU in BIOS (changed from SG to PEG), installed AMD drivers, loaded up a stress test and card had 75C load temps which are incredible since 85C+ were common on the old 6990m. Played some BF4 and got great fps at 40+ on ultra.

Yesterday (5 hours of proper gaming later and about 2 days later) the new card decided to crap out too, laptop just randomly crashed with horizontal lines, got 8 beeps at POST indicating no GPU detected, reset CMOS battery (this carry on is getting old now ) GPU was detected this time as ATI GFX in BIOS, had to resort back to Intel Integrated Graphics as drivers would no longer install, and yes tried the safe mode driver sweeper etc process to get rid of old ones.

Has anyone heard of this type of thing before? A card frying after working perfectly through a 1hr stress test and then a few hours of gaming and then just suddenly going? A card detected in device manager but failing to boot when PEG is selected in BIOS? I am stumped and this is my work computer as well and Intel Graphics really don't cut it for CAD lol.

Thanks for reading any comments / advice would be appreciated.............

 

Well, it seems like 7970M doing no better than the 6970M\6990M series in terms of reliability, I hoped for the better actually. It basically means that most of these cards have some kind of production flaws or\and have been not professionally repaired by chinese guys. I suggest opening cases for both cards (if you payed via paypal), at least the last one is obviously not your fault.

As for the first card, in fact if it was working properly - it should have triggered power off on critical temperature, as it didn't - it looks like it burned under critical temp, that mean it is a design flaw.

Just stay away from ATi cards for the moment (until there are any statistic failure rate info on their latest chips), you would be much safer with Kepler | Maxwell based Nvidia videocard for the moment, like 680M GTX \ 780M GTX or 970M\980M ( I would suggest skipping 880M GTX).
Or if you really do a lot of CAD - go for Quadro K3000M\K3100M\K4000M\K4100M etc.

 

Every time I see 7970M pop up on the forum, the post almost always says it died. Man AMD made a mess...

 

It's not like we didn't have the same behavior for the other brand with quite a handful of their products. As for the more recent 680m+, just give them some time, they started popping here and there. I mostly blame the unleaded solder that was the main reason for the whole 8xxx failures, they kinda managed things afterwards. That and 100W+ GPUs in a laptop. It's a spare part, if it fails, get another one. When you blow the engine in your car you don't go to cry over your, at the time, prefered brand, no? And no, I'm not made of money either.

 

Thanks to everyone for your replys. Not giving up yet, sent the second one back to china as faulty, hopefully will be able to get a replacement for it. Yes id love to make the transition to nVidia but that requires another heatsink which is an extra $150 on top of that, again from china. I envy you guys in the US who are able to buy these parts on domestic shipping, and deal with sellers who speak your language, I have to try to explain to a guy on the other side of the world who barely speaks English why he should replace my card...........

 

If you have the standard 100w heatsink it should work for either Nvidia or AMD.

 

are you sure? if the guy in china decides it is faulty, i might ask for 680m instead, can someone confirm?

 

@Mitch8198 - I'm with you, all the good AND cheap stuff are stateside. You can get a bit cheaper from China, but boy, does it come with a head-ache. I wont be surprised if they sent you faulty in first place, it's like 30% (faulty) rate for me. I have quite a few horror stories myself, and tons more around the web. Sadly, most US sellers don't ship international and the Chinese are getting more arrogant each passing day. Why shouldn't they, the whole world is made there.

Your heat-sink should be fine, just ask for a bracket, since I'm not sure that your would fit.

 

There is probably black tape on the bottom of your heatsink. You just pull off the tape and it will work with Nvidia cards. You can't interchange AMD and Nvidia x-brackets though.

 

I had this happen in my R3. Still have the laptop, and I had a brand new 580M to replace the somewhat aged 6970M a little over a year ago. The issue I faced was that the laptop WOULD NOT BOOT with the GPU installed. Take it out though and it'd boot like nothing had happened.

Still does it to this day too, nothing I do will work.

I was told it might be a bad southbridge on the motherboard, which would make sense given that the 6970M didn't work either BUT the 580M and 6970M worked in other machines without any issues!


Perhaps check the grounding on the heatsink and possibly consider replacing the motherboard.