Alienware 17 60Hz GTX980m upgrade

Done

 

Unless you arent online i dont see you online. But i did accept

 

I managed to get my 7970m to 1020mhz after bios flash. God i cant wait for 970m..

 

@DARCODER , And all members who upgraded to 980M please help me with this.

Hi,

I just got my 980M from Woodzstack for the upgrade but seems like I'm having an issue with thee GPU fit.The thing was the gap was too significant the screws from heatsink won't touch at all, have to pressure them too much which would definitely break something...Or is that supposed to be that tight

I don't have the Eurocom's extra thermal pad copper sheet for the bottom part for VRMs, but when I try with using a couple of Thermal pads only, I know useless without a heat dissapiation unit, but I tried without them also but the heatsink won't get fixed so I had to give that a shot. I noticed that the X Brace which comes with Dell cards have the 4 screw-in has more height than the 980M Clevo unit. Should I get that thermal enhancer pad from Eurocom to finish the job ?

 

i don't think there's a problem with the fit, there's just nothing on the underside to help give a stable base for you to screw in the heatsink. even on my 780m it came with some thermal padding on the underside and it fit fine.

 

Oh I see but I cannot manage to finish the job, Here please have a look at the snaps (3rd Trial today..Notice the gap there)
https://imgur.com/a/lI1kA

 

it looks ok to me. you do need to push down on those legs when screwing in but so far it looks fine.

 

I thought the same but couldn't pull this off, felt it was too much since it needed much pressure to make the screw touch that hole atleast to drive in. How about the backside ? I'm guessing what you had, the Eurocom special pad or the same setup ?

On a side note

Spoiler: Windows detects it as basic ms display adapter...

Good thing is It's detected as microsoft basic display adapter I'm good then. So 860M was really dead afterall (Noticed a chipoff at the edge of the core, maybe high pressure, but the repaste job was done >10 months ago, will be careful)

 

Nope, no special pad. I wanted one but never got it. the padding that was already there was enough. so long as the holes in the legs line up with the ones in the card you should be fine. that's why you need the padding on the underside.

just make sure to do a quarter to half turn on the screws so they're not overtight.

 

I take a quick look and for me it seem to not be the exact xbracket from dell. Here a picture how a xB dell look:

http://g02.a.alicdn.com/kf/HTB1jd3M...t-MXM-3-0b-n-Vidia-GeForce-675M-680m-580m.jpg

To make it work you right you will put lot of pressure. I allready try it once and it is working but lot of pressure. If there is too much pressure, the badest will be your heatsink core screw(s) breaking. Or buy/exchange for the right.

 

Also put some big pads down the gpu for both ram. For example, the china guy where i baught my first 980m told me the vram down the gpu died because of heat.

 

I've got my old 860M thermal pads, Should I go with them Or order some ?

 

both: use the old ones but order some in case

 

I'm checking with Fujipoly and I'll consider either them or go for the TEP...

 

how much is the shipping?

 

Well I checked for the large sheet first the price was high
The small ones which we need start at 8$ for the Mosfet block - 10$ for the square block both are at 11w/mK & there is 17w/mK from this thread those are not needed imo..

Also check this thread on backside thermal padding & Xbrace. So I'm guessing the Eurocom's TEP would be better since It has the heat dissapiation copper sheet to spread heat rather than just slapping those thermal pads onto the GPU without any heat transfer (crude heattransfer we can say since those pads get in contact with the chassis but the plastic film will insulate the heat again..)

EDIT : Shipping cost - USPS Priority mail (2-3days) is 8.70USD, TEP may cost from 35CAD I guess + shipping from woodzstack

EDIT 2 : TEP is a metal backplate. The TEP I was saying all long was actually Eurocom's thermal pads for backside, my bad.

 

You can use old, i always used older one. i replace only if it's broken. You can also wash them with a bit of water but be sure to clean them very well after ( to be sure they are dry!)

Clevo bracket and dell are a bit different. You might not have notice it, but clevo fixation is smaller than dell: that's why you feel you need to put a lot of pressure. I have both of them here.
==> if you put a dell xbracket in clevo, you loose lot of pressure= higher temp. If you put a xbracket clevo in dell system, you might risk to break heatsink ( it happen in my m17x r3 mod heatsink but never in newer heatsink). And screws for both of them are different. Dell screws are thicker than clevo.

 

Thermal pads supplied original with your gpu, are of good quality. You dont need to buy new one.

 

Installed the GPU and booted into windows and detects as Windows adapter, I disabled it and extracted these files and ran that modme.bat as administrator and this shows up for a millisec and goes, I tried to get a screenshot of it..
https://imgur.com/NY5oQi9

Edit : Ok I avoided the bat file and issued a command but says the PCI subsystem ID mismatch, should I proceed ?
https://imgur.com/XmmLfH7

Edit 2 : Many thanks to @woodzstack, @thegh0sts & @DARCODER, Prema and J95 without whom I would never able to finish this upgrade.

With fans Maxed out & chrome opened, the temps are - 32C GPU and 37-45C CPU. GC Extreme.
With fans maxed it runs way cooler than human body !! Both are under 35C

Now gotta try the games and all benching stuff !

A quick run of BO3 maxed out - CPU at max 64C and GPU is 72C


EDIT 3 : With normal fans on Auto some webbrowsing under 50C (45-47C both), And the card is running without any throttle with Prema vBIOS.

 

→ → http://www.guru3d.com/files-details/nvidia-inspector-download.html ← ← . Congrats with new hardware

 

I was getting the same temps at idle.

Sent from my SM-T560NU using Tapatalk

 

Some observations,
Had to install the 11.4CFM fan , was running remember me game and the temps went upto 70C I was shocked, So I removed the old 1.74CFM and installed the M17x R4 fan and the temps are now on at 60s, blackops 3 first run was nothing at all to consider, now the temps while in real mission are varying with the new fan at <71C which was last time's maximum at the start of game at 20-30secs run only...

Did notice the sound more with the new fan though Well one more thing I've observed is the idling of the 980M, it idles at the auto fans from 44-50 when the 50C reaches the fans spins for a few secs at a bit high RPM and then drop to 43-44 again, happens with CPU fan too. I guess the GPU fan is noticeable now because of the higher CFM and the cycles are lesser, takes time to ramp up the temps automatically for the GPU, then the cycle repeats.


The dell engineers could not make a proper fan profile in the sBIOS EC code. Just a mild run wouldn't be much to hear if they wanted a beast run silently....meh the idiots are prevailing !!

 

Dell started to destroy Alienware already with the 2013 models before everything became BGA. Worse fans and fan profiles than previous models.

 

The problem is nearly 95% of the laptops (not just gaming) released for the past 3 years are all soldered on cpus, except for the super high end clevo models and a few more, eg. Alienware 18, it seems like soldered processors are the norm currently.

 

Yet another reason to build an gaming rig

 

Guys I'm getting this new issue,
Whenever the GPU reaches up 85C, instant black-screen have to force shutdown, Happened once with BO3 and I re-pasted 1 week back, didn't get to play more and now I tried Technomancer same issue re-pasted just now and again same issue, I never had this type of issue before with Bo3 I believe I used to get to high 80s rarely (don't remember the exact numbers) before but never got blackscreen and had to shutdown. To mention today I forgot to use the HWInfo to max fanspeed on GPU when CPU was under full load temps went to 85C but that time Laptop was working fine and I had to immediately switch on the Fans though...

One thing to mention, Since I didn't use the alcohol prep pads to clean some fine threads
(from bounty) got stuck in the black tape which sits outside of the GPU core cleaned most of the part but I believe this is not the issue but something to do with the GPU or Cooling..

Also to mention the some thermal pads are blue ones (the extra ones that I got with @woodzstack kit which are not OEM quality only 1 pad applied tho on the back of vRAM) is the vRAM overheating causing the GPU to turn off, I don't think this is the case here though...Just throw in your opinions, I'm all ears Is the GPU at fault or what ? This is driving me nuts...

Please help. Thanks !

Edit : @thegh0sts , any idea here bruh ?
Edit 2 : Tried Heaven and pushed the card to limits waited for 84C and then instant blackscreen, I'm sure the GPU temp went to 85C point.

And as a sidenote my U3+ mod is still WIP needed 1Y splitter, coming from China so takes time, Although stock cooling shutdown is too much of an issue. The powerdraw is 114W on the GPU when running technomancer maxedout with CPU at 25W.
Tested with 362.00 same issue.

 

IMO 80c is out of the comfort-zone for Maxwell seems a bit high, it also depends on ambient temps.
My 980M used to black screen sometimes with temps @73c , after ~30min- 1 hour of gaming with an OC. Repasted with Mx-4, IC Diamond, nothing seemed to work. Then I remembered a thread where someone mentioned that MOSFETs get hot and black screen with Maxwell.
I repasted with GC extreme, Applied Fujipoly thermal pads 17.0 W/mK. The card never reaches 67c @ 1300/2905 1.075v running heavily modded Fallout 4, GTA, Heaven bench, Valley bench, firestrike and 3DMark 11.

 

Update -
A clumsy job - I opened the laptop, removed the MXM, removed the black tape on the GPU now but before that sprayed too much alcohol and it seeped through went into that plastic cover that stays on the mobo had to open that , that plastic cover was hard to remove opened a bit blew some air to dry it, seemed like it had some electronics friendly gum which got mixed with the alcohol and the TIM (lil of it) thinking alcohol might stay there and short but it looks a bit not clean now , Damn I wish I had read @Solo wing post earlier, re-assembled the machine working, then had a heavy breath.

Ordering the alcohol prep pads and arcticlean kit, fyuck that 70% IPA spray and bounty they scratched the GPU / CPU cores. So as people said the GTX 980M is indeed the scrapyard of the GM204 chips.

Ill check for the fujipoly too then the thermal pads look worn out, @Solo wing , could you link me which size should I get. Thank you very much for the insight !!

EDIT : is this okay ? or have to order a full sheet ?

 

-double post-

 

-double post-

 

I bought this one. 60x50x1.0 It was enough for the front vram,vrm(mosfet,inductors.....)

 

Oh okay, tq. And the back does the vRAM stack don't they need ?

 

I put stock dell pads, the white 0.5-1.0mm ones. Also I still have the large silicon pad from my old 680M, it goes under the GPU and it's quite big, bigger than 60x50.

 

does anybody have the vbios file for the GTX 980m on an Alienware 17 (late 2013 build, 120 Hz, but I guess that doesn't matter)?

 

https://biosmods.wordpress.com/gtx9/