M6700 upgrade from FirePro to HD 7970M not booting.

Hello,

I've bought a eurocom 7970M from a local friend, Someone I trust and therefor I assume the card is functioning.

Now I've simply put in the 7970 but my laptop will not boot, The fans go on for about a second and that's it. The LCD screen just stays black, the backlight doesn't even turn on. Once I take out the card and boot on IGP graphics it just boots as normal. I'm kind of dead in the water here since I've traded my old AMD FirePro 6000M for the HD7970M and I've run out of idea's of what the problem could be.

I'm running the latest A08 bios, Downloaded from DELL itself.
I've tried booting using external display's such as VGA,HDMI and Displayport, all no avail.
I've reset the CMOS multiple times using multiple ways.

I've made pictures of my card, Maybe somehow they're usefull



I've understood that the vBios on the 7970 is only for BTO/Clevo laptops or something, Yet other people say it should just be plug and play.

So yeah, Here I am now with a semi-dead M6700. I'm really hoping someone else here will be able to help me out with this issue.

 

The 7970m is a faulty card, get the refund or replacement. I have experienced the exactly same symptom 2 times, and in both scenario the card was dead. 680m. overheating (heatsink not installed properly).

It is plug and play. You made really big mistake for selling the M6000 before 7970m is tested running fine, at least for 1 month.

 

That card from what I see is a Clevo. The X-Bracket is not good for your system. Use the one that came with your M6000. The thread hight is the problem causing your GPU to not touch the heatsink and thus overheating and not booting the notebook.

 

Not really. Even without heatsink the GPU will still boot the notebook (show the post). In this case the card is already dead.

 

I think you already fried your card. You booted basically without a gpu heatsinks. If you can get your hands on another one make sure you use the bracket from the firepro card.

Sent from my GT-N7100 using Tapatalk 2

 

With the look of the paste you have on gpu chip there was contact between cooling and chip itself and you had some cooling which should have been sufficient for a boot. Even if the heatsink was not properly attached card shouldnt die from overheating in a one boot sequence. Thermal protection for shutdown should have prevent that. Amd introduced that long time ago. Last chip they had without thermal protection was Athlon with Thunderbird core.

Over the last few years working and repairing desktops and notebooks i've get my share of users who did not properly attach heatsinks to their components and even tho temps were crazy nothing got "burned", nothing had problems except nvidia 7xxx and 8xxx series gpu. Where gpu didnt die due to overheating but rather to very bad gpu solder quality. Reheating with 200-250w bulb or "baking" it in oven could have fixed the issue in some cases

Long story short, my bet is on receiving DOA card, if you didnt damage gpu chip itself.

 

Aye, pretty much that's what happens.
I did have some scenarios with trying to boot a 7970m and even though contact appeared to be just fine, the card refused to turn off until every IC and mosfet was properly covered and cooled. Very sensitive card. I would suggest trying to assemble it a couple more times, and make sure you have nVidia Optimus disabled in the dell bios. Absolutely make sure that you cover all import areas.
This means in addition to what you have covered in your picture, make sure that the vRAM is properly cooled, that the inductor to the right (big silvery cube above your right-side vertical vRAM) is also cooled, and that the FET's (three black IC's located to the right of that inductor) are also properly cooled.
If not, try and boot the card in another compatible system, although it could already be dead.