Just bought a 7970m off ebay

Dell AMD 7970M 2GB GDDR5 for M17x R4 | eBay


Can someone who has done the gou swap let me know if everything I need (save for tim and pads) is included? Also, I see that it has been vbios oc, im worried that it will throttle as a result in the m15x. Can someone just give me a "putting a 7970m in your m15x for dummies" as far as vbios flashing and things like that? I am fine working with desktops, but notebooks are so small.... Thanks!

 

General guide to 7970m in M15x http://forum.techinferno.com/alienware-m15x/1731-[guide]-m15x-amd-7970m.html

How to mod the BIOS [GUIDE] AMD VBIOS flashing
Make a Boot-USB, put ATI-Flash on it along with the bios you want to flash onto the card. Follow the commands but include the name of your BIOS when running the command.
If you forget then type dir in dos to list all files present.

Here are a bunch of modified BIOS' Downloads - You can download for free but donations are suggested.

Also check the link in my sig for a video guide on upgrading gpu.

Best of luck.

 

Thanks for gathering all of that for me! +rep for you obviously

 

Just got home from the UK. 7970m installed with all the tips you gave me. Work like a dream. Idling at ~35c, load ~65c. couldnt be happier!

 

Glad its working well

 

I'm in a similiar situation, but with a few issues... Over the past few days I've read just about everything on the forums related to "7970m and M15x" to no avail... hopefully I can find some help! (imminent wall of text)

I ordered a 7970m from ebay, and got some Antec 7 Thermal paste along with some good thermal pads (The brand escapes me atm). After a smooth installation (properly applied thermal paste/pads, was super careful not to muck up my card), I booted it up and tested out The Witcher 2 on it. My gpu temp skyrocketed to 83 C and soon after I hit the auto-shut down temperature, causing my laptop to shut off. After some underclock testing and switching to a dozen different drivers (to no avail), I remembered a discussion on the forums that the GPU die doesn't make contact with the Heatsink if you upgrade from certain cards (in my case, a 5xxx series AMD card), since the x-bracket screw holes sit higher on the 7970m. So I opened up my laptop and lo-and-behold the heatsink and GPU core weren't really touching. I read on some forums that people have used copper shims and the like to ensure good contact. I didn't have one available, but improvised using a copper penny covered in thermal paste to ensure a good seal. It worked like a charm, my GPU temps went from 85C to lower than 70C when running the game, and idled at 35 C.

Excited but exhausted (I had been working on it for hours), I shut down my laptop and went to bed. I woke up that morning, started to download Planetside 2, went to work, and came back home. I started up Planetside 2 and carefully monitored my temperatures while playing. I peaked around 72C after around an hour of playing (awesome), however soon after my computer completely locked up, and displayed light blue vertical bars across my whole screen (as if my card failed or something). Not only did it crash my game (obviously), but my entire laptop was completely unresponsive (my skype call with my friend even stopped). I hard-reset it, let disk check run just in case, and booted up my laptop. Thinking it was some modded drivers I had installed from the day before, I uninstalled them/cleared them off my system and updated my drivers to the 13.6 beta drivers from AMD's site (I even took the time to allow the default windows drivers to install first, as some people have discovered that helped).

I rebooted as required, started up Planetside 2 again, and I got a light blue screen with vertical bars again, this time in an even shorter time (~15 minutes). After shutting down again and after many attempts to load into the game, I tried TF2 just to see if Planetside 2 was at fault. I got the same screen again (at the server connecting screen!) and my computer locked up once more... thinking I might've melted something on my card I opened up my laptop again to check my card out, it looked completely fine. I noticed some of the thermal pads weren't making good contact with the heatsink so I put on some more copper pennies on top of the pads (with more thermal paste, and took my time doing it, making sure not to bridge any connections), and booted up. I loaded into tf2, got the same screen and shut down. This time, when I tried to load up, the numLock and scrollLock lights were flashing while the caps lock key was solid. Apparently this indicates a problem with the LCD screen or the video card... Frustrated, I opened up my laptop again, removed all pennies/thermal paste (except the penny on the GPU die, since that actually helped), and my computer booted up again.

But, I tried to load up TF2 again, this time underclocking both my CPU and GPU via stealth mode to see if that would change anything, and I still got the screen with light blue bars and my computer required a hard reboot.

The only possible things I could think that could be the case is that I need to get longer screws instead of using a penny-shim on the gpu die to ensure proper contact with the die and the pads... because maybe the thermal pads on the rest of the GPU are too far away from the heatsink, and that's causing the VRAM to overheat on the card. Or the 13.6 beta drivers are somehow not working out for me.

Or I could've bought a bad card. In which case I can return it...

Do you guys have any thoughts/opinions? Should I just get new screws and see if that fixes it? Should I scrap that idea altogether and try something else? Do you guys know why I got that weird screen/what it means?

Computer specs:

Intel i7 740 QM
AMD Radeon 7970m
8gb 1333 DDR3 RAM
Bios A09 (up to date)

Also I wouldn't think it's power consumption, as I'm only running a non-clocked 740 QM... So I'm completely out of ideas...

Any and all help is appreciated!

 

Using a penny on the die sounds like a really terrible idea! I would take it off immediately and just prey you didn't do any damage to the die itself.
Try to find some longer screws, removing the C-clip screws if they are fitted on the heatsink. This should resolve temperature issues. Not sure what is causing the blue bars unless you were shorting something on the card.

 

Off the back of this post to save making a new thread do you think that this is a fair price for a 7970?

Alienware AMD HD DELL 7970M 2GB GDDR5 -for M15x, M17x , M18x ,upgrade monkey | eBay

I know you can get better deals on NBR marketplace but i generally seem to miss them.

I already have the X bracket needed.

 

Reasonable price but I've heard some people have had problems with upgrademonkey so buy with caution.
I'd much rather buy from a trusted member here on the forum.

 

Oh really? I bought my ram from them 6 months ish ago but thats all and it was quick and hastle free.
Maybe i should wait then as a gpu is more of an investment.

 

http://forum.notebookreview.com/computer-components/719296-fs-dell-amd-radeon-hd-7970m.html
http://forum.notebookreview.com/com...md-radeon-hd7970m-dell-amd-firepro-m8900.html

One UK 7970m and one US. Also Pau1ow has 2 posts listing 7970ms but he hasn't responded recently so its hard to say if they are still available.
Maybe send him a PM?

 

Thanks, Pau1ow has sold his, spoke to him last week. Have PM'd the other UK one though

 

It was meant to be a temporary solution, and I triple-checked to make sure no actual copper was contacting the die, only thermal paste. Also if there was something shorting out on my card I'd be willing to bet I would see melting/burning somewhere, luckily I carefully examined the card and that's not the case. Also I should mention the laptop is completely fine just browsing the internet, or doing low-demanding stuff, it's just when I try to play a game I crash.

But you could definitely be right, I'll head in to a hardware store today and get some screws... wish me luck, and I'll update when I get back.

 

Good luck! I posted a picture here of the screws I used and the original C-clip screws that were too short.

 

Just remove c clips from heat sink and then use screws from name plate. This is the way I am rolling and it works good, temps only go above 75 under furmark and it takes about 12 mins of furmark to make it touch 85 and then shutdown.

Bf3/civ3/serioussam3 makes it stay around 75

 

^^^ The man's a genius! ^^^

 

Alrighty, well after a long afternoon of many stores and trying to find good screws, I saw that you could use the nameplate screws (thanks vs3074!), and gave those a shot. They workd perfectly, and I re-applied thermal paste and redid all of my thermal padding just in case. I also applied a small amount of thermal paste on the tops of the pads to ensure they made good contact (A couple of the pads were just barely touching the heatsink).

It looked extremely promising, but When I went in game again, I got vertical lines across my screen and my computer crashed once again... so that must mean it's something with the card. I think I'm going to return it and get a new one, probably from the NBR market.

If anyone has some more advice/ideas let me know!

 

Replace the pads, since you put thermal paste on them, it can mess with there conductivity. 7970m is super sensitive to static, just adding pads at the wrong place can also cause static and messed up video output. Quite frankly I am amazed at how sensitive 7970m is compared to 5870m.

I had the experience where I added old thermal pads underneath the card to make thermal conductivity better but one of the pads had a small (0.5mm thermal paste spot) and it gave video equal to faulty card, removed the thermal pad and all working perfectly fine.

Now only if I can get enough time to open my m15x again and rearrange thermal pads so 7970m starts working again (garbage video output right now with reddish static).