Windows 10 and nVidia 980M GTX upgrades for Alienware (M17X-R4, without Optimus: Working Great!)

I was thinking the same thing.

 

@woodzstack

Hey man

Sorry took a while for my laptop to do the system restore and finally boot... Here are a few screenshots of the issues I told you about by PM on eBay...
Really hope you can help me with this.... I can provide teamviewer if you want to take a closer look at things...

 

whats your hardware ID ?
Did you read the original post of this thread ?

 

PCI\VEN_10DE&DEV_13D7&SUBSYS_05AA1028&REV_A1

Well I read your original post the day I received the 980m from you and it was a great help getting it to work. I've been using it since November without a single problem... Aside of never actually getting it to work on Win10 so I just downgraded to Win8.1 and have been gaming on it ever since. Never OC'ed it aside from a single time when I was benching it the very first day I received it (used NVinspector). I never installed any OC tools like precision or any others...

Edit: Raidriar on the AW17 thread is suggesting I try installing my 880m again to at least check if the MXM slot is dead/damaged on my board. But don't you think that if it was the case, the card wouldn't get detected at all? To be honest, I am completely out of options so I am going to try installing the 880 anyway...

ust installed the 880m. Bad feeling about this already. Same error in drv mgr already... Installing driver and rebooting now...


Sent from my iPhone using Tapatalk

 

I think its the driver, because of the bad history nvidia has had in the last 12 months. I think its nvidia even without evidence at this point.

try this : http://downloads.eurocom.com/suppor...ptop/Alienware/17_R5_05AA1028_9xxM_358.50.zip

see if it installs after disabling DSE, and running it as an administrator.

I would also update ALL your hardware and windows updates before continuing at all. I've seen that be an issue about a hundred times, but only with windows 8 and windows 10. There are small hotfixes and updates that affect firmware and hardware and how it works, in those updates sometimes that fix everything. The issue is, windows likes to update now, without telling you they're removing or disabling previous updates and hotfixes, and until you reboot, finish updating, sometimes things just do not work properly, maybe fans stop responding, or system can not detect hardware properly.

SO, start there first. Remove ALL evidence of any videocards, using DDU in safe more, then continue with installing other hardware and updates. once your 100% done and rebooted, reboot again, disable DSE and install that driver, or try too. (Sometimes the drivers on that page are messed up and don't work properly, if so let me know too, too many to keep up todate all the time)

**added some extra content to Original Post, vbioses, a PDF, some extra specific instructions, etc..****

 

Oh well... Driver installed fine, 880m recognized... BUT gpuz render test shows nothing. Nada. Downloading unigine valley to test it. If it also fails then I guess it's a mobo issue... right?

Edit: false alarm, just tried 3dmark and fallout. Both working fine... No artifacts nothing.... So it seems my 980m is defective.... Will do some more tests tomorrow before having the warranty talk... I do get 90 days warranty right?

 

@woodzstack

Yeah the 880m is working fine... Gotta love gelid extreme... Max temp during firestrike=71... Before I replaced it with the 980m I used to get highs of 85... Never repasted it though. Anyway it is midnight and i need to wake up at 6am for work... Will test the 980m again tomorrow...

But damn is the 980m a better card... I got 9000 on firestrike with it while I just got 5400 with the 880m...


Sent from my iPhone using Tapatalk

 

85C is abit higher then you should get on the 980M. What heatsink are you using ? The one you got for the 880M that came with the laptop ? There has got to be some reason why your 20C hotter then everyone else under load. You probably thought that was fine because your 880M gets hot like that, but the 980M only really gets that hot if its not being cooled properly. If I had to guess your using a NV100 heatsink, right ? I've even seen people use a heatsink that was bent and they didn't see it. Think you could take a picture and show me, please, I've seen all the heatsinks, I know them well enough, and say hypothetically its the heatsink, then even if you burnt out the 980M replacing it would do no good, right.

In every picture you've shown me, you were missing updates and most your hardware was not installed, this is another thing that could cause poor performance, always make sure all hardware and all windows updates are done, before touching an aftermarket videocard.

 

yeah, make sure all drivers are installed before troubleshooting.

why HWINFO32? are you not running a 64-bit OS?

if you're running a 32-bit OS you should go up to the 64-bit version.

if you're running a 64-bit OS use the hwinfo64....if this is the case then you're really confusing ppl.

 

This is an AMD heatsink, you can see where there used to be the black tape. Its different, it forms a diamond because the difference with the GPU die on the AMD is oriented different. Remove the tape and it works on a nVidia card.




Now the NV100 heatsinks for Alienware laptops have this metal part in the middle near the mosfets and capacitors above the RAM...

The only way to use some of the nvidia heatsinks is to make sure that middle part is not preventing a proper fit, sometimes it sticks in the way and overlaps the capacitors, and it only marginally looks like a proper fit, but in reality, its not actually levelled on the GPU properly nor the rest of the card, this is what I've seen people do, and they get about 85C - on any other card, you'd burn it out almost instantly with running a benchmark tool. The 980M however do not actually get very hot when the ASIC quality is high. Ones with 60% ASIC still act like a 680M / 780M did or 880M.. they get hot and have no thermal headroom to overclock etc.. even getting 9000 on Firestrike would be difficult because it wouldn't keep up Turbo.

Anyways, on the chance this could be the issue, let me know.



this is perfect fit (this is your heatsink picture you sent me.)

So lets rule that out. I think it might be your fans not working, because it shouldn't get hot. Just doesn't make sense, to me. I would start an RMA asap, could be something is degrading, and by the time you get a chance to send the card in, at least its not entirely broken.
SO, do you use any program that might have access to your fans ? I know using those actually screw up fan control sometimes, like using evga precision/MSI afterburner or especially - hwinfo64 can actually force you to manage your fans manually. (I forget the circumstances mind you..) sometimes a windows update stops them from working until you reboot, when it does small updates. If its getting hot - to the point its crashing from being too hot after extended play, as you have described, the fans should be going max speed and sound like a jet about to take off in on your desk. They'd be loud enough to distract you from playing even if you used a headset.

I also looked into the 880M temps, 85C is too hot even for that. shouldn't be able to hit that temperature easily, and not instantly or right away, you should be mid 70's. on the 880M and mid 60's on the 980M. (Unless of course its 35C in your house or hotter.)

actually looking at the heatsink picture you sent me, a few thoughts to help you. you have about 2 x as much thermal paste as you should, and you can see that the laptop heatsink was not levelled on the GPU properly, all the paste is squished on one side more then the other. When your tightening the screws, make sure to do it evennly and not to put too much pressure, this might help temps.. might help alot. You should put thermal paste so evenenly, so thinly, that is semi translucent and see through - like how it looks on the heatsink part in that last picture, where it is running thin. Thats perfect, but its tricky to get right, but learning this now, will help you forever. I'm not saying this is the issue either, but I am curious how much of a difference it will make. you might have to try 2-3 times. When I used Gelid extreme on a 980M for an R5, i get like 58-61C under load when gaming for 1 hour, like playing the witcher 3 on an extrernal 4K monitor on medium settings... I would get about 66-67C with ICD7 (its soo friggin thick..), TX-4 gets me about 60-63C

Also, some of the components seem not to be touching the thermal pads, like the green / yellow limish one pad seems untouched. maybe use a thin thermal pad that I provided, and just see if it helps make contact. (should be 0.5mm thick - that one), and to support its not touching that component, you can also see on the GPU that is the side that is thicker in the thermal paste as well. Could be, the heatsink is ever so slightly bent on that corner, or it was just overtightened on the opposite direction.

Anyways, let us know about the previous things mentioned, about software first, and then once things are A-Ok to go switch to the 980M repaste/cleanup , run DDU before switching though (i.e after DDUY thats when you remove the 880M)


when it comes to the screws, you don't force them all the way. Just a quarter turn on the one labelled 1 then move onto 2 and then 3 etc.. and do this until each hand put enough pressure to start to bend the metal arm that is attached to the heatsink, there should still be enough room to screw the screws in about 2-3mm. anymore then that and its overtightening, and most likely causing microfractures on the GPU die.

You can clean the GPU die up and take a look, if any of it is cracked or chipped off, thats considered physical damage. From how the thermal paste was pressed thin, I can assume the heatsink was overtightened, but if the card still works, maybe it hadn't been broken yet, cause of it. Thus I am not saying yet that this could be the issue either. But a close inspection with a magnifying glass and cleaning the surface up on the GPU should show you if there is damage.


if you look closely on the RAM itself, some of it has no smudges in it, and its consistent with the corner that might not have been making proper contact. Like the one riiiight in the corner between both lines/sets of RAM. I hope this stuff helps figure this all out, and I hope the card is not damaged, but it does seem like there are small indications of not being levelled out evenly.
The temps were the first indication. The indentations on the thermal pads are probably from the previous videocard. You could always throw some very cheap thermal paste over the RAM and components , evenly spread out, and spread thin, to see what kind of contact it makes with the thermal pads.

 

@woodzstack
@thegh0sts

Guys, I am not trying to confuse anyone. I don't know if you realize what I did in a matter of 3 hours last night.
I actually fomatted THREE times my OS drive, did THREE clean installs of Windows 8.1.
I reseated, repasted the 980m. Didn't work.
I replaced it with the 880m. Worked after using DDU twice and finally getting it to work on latest official drivers from Nvidia.

The 85 Degrees I used to get were on the 880m before I bought the 980m as I mentioned in my thread. I never got more than 64-65 degrees with the 980m. I got ZERO issues with the 980m prior to last night. The ONLY thing I did recently was switch from Optimus to Dedicated 3 days ago. For 3 days my laptop worked fine without a single indication that something was going to fail.

The reasons you see so many uninstalled devices in my screenshot is that I was sick and tired of installing manually all the Dell drivers after THREE formats, only to end up with a BSOD again and having to format again.
I will do things from SCRATCH agian tonight. Full format, full reinstall of Windows. Install all drivers, all windows update. Heck, I will run DDU in safe mode even before installing a single Nvidia driver on this clean install of Windows. Then and only then, I will attempt installing the 980m driver again...

I work in IT (ERP project manager), and I manage FOR FUN, my company's network infrastructure, 5 servers, Cisco firewall and PBX. I only mention this so you don't mistake me for someone who's clueless when it comes to computers. My OS and every single application on it is always up to date. All drivers are up to date. I have ZERO malware on my PC. I run CCleaner practically every 1/2 hour by principle as I despise the concept of cookies and any temp file being written on my drives without my consent. I have NEVER been infected with a virus for at least the past ten years.
I know you mean well by being thorough in your suggestions but believe me when I say that the problem I am facing with the 980m is NOT straightforward and not due to any negligence on my part. I do live in a hot country but it's winter right now and temps don't exceed low 20's and I use a cooling pad anyway for my laptop. Idle temps for the 980m are 36-37 and as I mentioned, they never exceeded mid 60's. The 880m after repasting with Gelid Extreme doesn't exceed low 70's.

Regarding the overtightening of screws, I had used such advice from someone when installing the 980m in the first place. I did inspect it thoroughly when reseating it and removing it. At a first glance nothing seems wrong. I don't have a magnifying lense but there is no visible damage as far I can see.

I will let you know tonight how things goes after I try reinstalling the 980m.

 

@nick81

what does that have to do with this situation ? We're trying to help. Your possibly skipping steps. Shouldn't be possible to fully install configure and update all windows updates and hardware in 3 hours, from scratch. Do it once, do it right. Theres no need to do it multiple times unless there is something your not telling us, that was causing an issue too.

CCleaner does not guarantee no malware, just saying. Now, I've asked before about the fans, I will iterate and ask again (thrice in the emails and once in this thread already) are they working properly ?
Did you ever use HWInfo and notice the fans stop working afterwards ?> Did you play with the profiles inside the bios when using an unlocked bios before...
If your sure its not the bios, reflash that, start fresh, then redo your settings. Then when in windows - update to the latest MEI Firmware too, can't be too safe right. Of course then do the OSD and AWCC applications too, not just all windows updates, without those programs sometimes the fans don't work properly. Don't know why, I might even be wrong, I'm no software engineer.

Check what vbios you have in GPU-Z, also while your there, check for me the ASIC read out, and check it a few times when your troubleshooting, if it crashes or gives you the black screen, I want to know if it reports getting lower, that's an indication of slowly burning out the cards from overheating. I've once had a card go from 74% down to 61% in a single day of pushing its limit, and running it hot - like in the high 80+ all day, this can happen when overvolting, so lets check that too, I want to see what the voltage reads on MSI After Burner.

Now, when you have MSI After burner installed, see if you can use that to control the fans and set default back. Then see if you can run manually again, and put it to max speed - and then see what happens when using the card for an hour, what are the temps, whats the voltage, whats the behavior we're getting, whats the vbios, whats the system bios, whats the ASIC and is it moving.

Also, I asked before if you could try using some paste to see if the RAM is touching and making proper contact or not. Could be the RAM and the card on the 980M is potentially slimmer then the card you had before, the 880M- this is why i usually sell my cards with thermal pads too, they're extra thin, just to make up the extra slack when you re-use the old thermal pads, and is why i tell people not to throw out the old ones. They're meant to go to gether.

Hope this helps. Could just be a thermal issue/cooling issue, or a vbios, or a slowly breaking card, but in trying to eliminate if the card is not broken, we need to try everything. Since the card was tested before being shipped, thoroughly, and you even had it working, before all the re-installs, its hard to say what is what. But - from anyone who knows the 900M series, they run cool. VERY cool. Not hotter then the 880M thats for sure. So, right there is an obvious sign.

 

@woodzstack

I only mentioned my profession and the field I work in, to indicate that I am not clueless when it comes to PC's and the potential problems one has to face when using them...
I know you're trying to help but I feel like we're turning in circles. You keep pointing out problems with my temperatures while I keep mentioning that it was probably a misunderstanding as the only temperature issues I had in the past were with my 880m. It was ONLY due to poor TIM as with Gelid Extreme I am seeing an idle temp of 32 degrees right now on my 880m.
With the 980m, I have NEVER faced any temp related issue. Not one artifact, not one black screen, not one BSOD. I don't OC, I dont mess with BIOS, with fan profiles. Both my CPU and GPU show more than decent temps thanks to my cooling pad and quality TIM used.
I am spending some time with my kids right now but as soon as I am free, I will disassemble my laptop for the 4th time in 48 hours and try installing the 980m... I will do my best to give you the readings you've asked me about. But like I said, I can only do that while the GPU shows up as a "basic display adapter". As soon as I install the driver and restart, I get a BSOD.

In the meantime I have a question for you. In HWInfo, after installing the driver and BEFORE rebooting, I can see a reading for the GPU's RAM. It shows 1 GB... Could this be an indicator that it's an issue with the RAM ?

 

Hi , Sucessfully Upgraded my Alienware m17x R4 to 980m which i bought from sir james or woodzsrtack666.....
heres the proof https://www.dropbox.com/s/hp3tpd84an505be/Untitled.png?dl=0
First i Got 8 Beep and no post .. then what i do is reinsert my old gpu which is 7970m then check if my lcd still working then it works and went to login windows ... and then i reinsert the 980m and now sucessfully booted and install the driver from sir james/woodsztack666 from his dropbox then device manager sucessfully detected the 980m ...

 

@nick81

(moved this here)

Okay, so, you have an AMAZING card, one I sent with an OC'd vbios, overvolted... hitting 85C is only normal because its about 30% overclocked. Its essentially faster then those 980 Desktop cards they just came out with. But, it needs better cooling.
You can easily sell that card for more then you paid for it, unless you ruin it. If you want a lower ASIC I will happily exchange it for one that is 70% and has a stock vbios. If you put it for sale on the forums here, theres about 5-6 people looking for a card with that potential.
So, I would say there is NOTHING at all wrong with your card, the fact remains, I might have given you something bigger and better then you can chew, thats all. but you asked for a good card too, if I recall, so....all is good, no ? Just flash it with a less crazy vbios.

Most people buying MXM cards are getting ones that are cheap and pulled from laptops, and being told they're new, and high quality. You will rarely if ever find one that good, in a laptop, no matter how much you pay.

The vbios I sent is OC'd because the cards is very very capable of handling a higher OC then normal. Just flash it down to a stock vbios, and temps should drop by 30C easily. mainly the voltage is supposed to be 0.87 when in load, your is idling at 1.06 basically 25% more, thats why the temps are so high...but the card dishes out very high performance. Very high.

The 8 beeps is caused by settings in the bios. Like having wrong UEFI settings then are needed, as an example, or like when hitting the switchable graphics Fn_Fx button and rebooting .... people get that 8 beeps ALL the time... its not as bad as it used to seem, when people didn't know why they were getting it.

 

@woodzstack

Hey man

I am going to try and clarify this one more time. I have NO issues whatsoever when it comes to temperatures with the 980m I got from you. Idle is at 34 and I have never seen the temps exceed 63-64.
Why would I flash it down to a stock vbios if I haven't faced a single issue with Prema's vbios ?

I apologize as well for not sending you the info you asked for last night as I honestly was too tired to attempt anything. The night before I had spent 4 hours on my laptop and I couldn't spend another consecutive night away from my wife and kids. An angry wife >>>>> dead gpu
Today's last day of the week so I'll definitely have time to do some tests on the 980m...

Thanks again for your assistance

 

am I getting you confused with someone else ? I thought you had temps high as 85+ and your system was shutting down after an hour of playing WoW or something, and even showed how it was idling at mid 50's celsius..

EDIT : after re-reading, I see you are saying you had issues only AFTER removing optimus/turning it off. Well when you do turn it off, the hardware ID changes, least it always did for me. That means your driver needs to change, if you used the ones I had made, since I only ever made them for a single hardware ID, so I would just check hardware ID and find the right driver and install it, when Optimus is gone.

Why did you use Optimus to begin with ? Although it does work, I have always warned against using it, for these sort of confusions. Always causes extra explaining and issues for no reason.
I get tons of emails and messages on various forums and threads all day, and places like ebay or for instance, have no way of benefiting anyone else, and they stop my messages after 250 a day or close to that, which I hit a lot, so when I try and help someone, I skim through their issues as best as I can, its more confusing when I get things I assumed were symptoms but they're explaining how it worked previously - my apologies for the confusion.

So check if your hardware ID changed. Then disable DSE (for good) and install the appropriate driver.

 

Lol! Yes... I don't play WoW... Do you mind going over my posts history? Maybe you can have some advice I can use tonight during my tests?


Sent from my iPhone using Tapatalk

 

was doing it as we type..

lets start fresh - what exactly happens when you turn on the PC when the 980m is installed.
Does it POST ? Can you get to bios ? Does it give 8 beeps ?

 

Yeah I was actually logging in to NBR from my PC to do just that

I will actually start from the very beginning to give you an exact idea of my setup and the upgrades I did on my AW17.

Note applicable to all steps below: I NEVER OC'ed my cards, never installed any OC utility aside from NVInspector.

1- I bought it with a 880m initially and a 1TB WD Blue hard drive cached by an 80Gb MSata SSD. No issues whatsoever aside from high temps in the 880m. I used to experience throttling so flashed it to SLV's vbios at the time and got a good cooling pad. Idle temps were at 38 degrees and I used to get max temps of low 80's. I didn't bother repasting as everyone online was telling me this was normal behaviour and that the 880m runs very very hot.
I have always used it in Optimus, never tried dedicated.

2- I updated to Windows 10 back on Build 9960 (tech preview). No issues whatsoever with any driver or software on any build till RTM.

3- I bought the 980m from you back in November. Installation under Windows 10 failed (in optimus. Didn't try dedicated). Impossible to install any modded drivers. Kept getting either graphic card not compatible or when manually installing from device manager, the 980m would show up, I would be able to use it for a few minutes, then it would switch back to "basic display adapter".

4- Reverted to Win8.1. I was able to install the 980m immediately. ZERO issues. Great temps : 35 idle and 66-67 under load with the IC Diamond TIM you provided.

5- I flashed it to Prema's 1.1.1. Zero issues gaming daily on it.

6- 5 days ago , I decide to give dedicated mode a try. I use the FN-F5 key, reboot. I use DDU to clean all drivers and install modded dedicated drivers. No problems there. I do a bit of benching, get great results (300 points more on Unigine Haven than on Optimus, better results in 3DMarks..) so I decide to stay on dedicated.

7- I am happy for 3 days until 2 days my laptop refuses to wake up from sleep. I do a first hard shut down. I get a black screen with a white square in the middle of the screen. 2nd hard shut down. I boot into Windows. Device manager shows the 980m with a warning sign. Error message about it being disabled.

8- I run DDU. Re-install modded dedicated driver again. Restart. I get BSOD before Windows login screen. (the page fault error). System restore FAILS. System refresh FAILS. I am forced to do full clean install of Windows 8.1.

9- I reinstall all drivers. I switch back to Optimus mode. I install modded optimus drivers. Restart. Same BSOD before Windows login...
This time system restore works.

10 - I diassemble my laptop, reseat and repaste the 980m. Reinstall drivers. BSOD again.

11- I diassemble. Replace with my old 880m. It works.

That's my situation right now. I sent you screenshots previously of my device manager error for the 980m (which can be ignored because I had the exact same message for the 880m BEFORE installing the nvidia official driver) and HWInfo.

One thing that you CAN'T see in my HWInfo screenshot is what happens when I install the modded drivers BEFORE rebooting. The main difference is that a value shows up in the memory field. It shows a value of 1GB which of course should be 8GB right? Does it mean that my GPU memory is damaged?

Forgot to add that while I was still testing the 980m, I was able to reflash it to Prema's 1.1.1 just in case the vbios has somehow been corrupted and I checked that it was well being displayed in the BIOS.

One peculiar thing that I noticed with my 880m BEFORE I purchased the 980m, was that sometime the GPU would show up randomly (happened 3-4 times) as a "basic display adapter" and that it wouldn't even be detected anymore in the BIOS. I would solve this by running DDU and reinstalling the official drivers. I never understood why this happened and why this could possibly affect the BIOS detection of the GPU...

 

okay, there is no issue getting it on windows 10 normally. You NEED dedicated mode for windows 10, Optimus won't work. There's a way around it, but its too much hassle getting it to work with every driver, because there's more then a single file to modify, and one of them keep changing.

So, your just on windows 8.1 then. Your not using the same driver are you ? For when its in Optimus and for when its in PEG/Dedicated, you have to choose one and then install the driver for that. Whenever you switch over, the drivers will be an issue again. Stick to PEG, keep it in dedicated, disable DSE install the driver depending on the hardware ID, you can also try a bunch of drivers here : http://downloads.eurocom.com/support/drivers/upgradeyourlaptop/Alienware/

I have to say try these steps, because it was not said exactly which driver you tried in each step you did, so I am assuming you used the same driver, and thus even though it was installed, its actually not compatible. Could probably avoid that by adding multiple Hardware IDs to the same driver's INF, so it works in different modes, but I did not do that, I would be going in circles.

Now, that advice is as if I ignored your issue but I didn't. I think your switching over, from optimus and dedicated all the time. If you can use an unlocked bios and set to PEG and keep it that way, it can avoid this issue.
Now, if your saying even if you do this, and were using ONLY the driver made for the hardware ID when it is in dedicated.. and still get BSOD, then I think you need to do a few preliminary steps.
Go to the first page, and go through all those steps, as if you did not know better. UEFI, all those settings, PEG mode, GPT Partition on the harddrive (note it can NOT have a NTFS partition on it, has to be 100%, this could be causing BSOD's when windows load, if you get that)

We're going to be resetting hidden menu's and things in the firmware, by doing all that redundant stuff that seems so silly that we might as well skip.. like holding hown a power button for 30 seconds when clearly 3 seconds does the trick if your resetting (or does it... honestly.. so many times, holding it down for 3 seconds... made me think it worked, and it didn't).

Also: Question : Are you using a 120hz monitor ? If you are this could change a few things, but following the steps on front page should work.

 

Hey man

Alright so I reinstalled the 980m.
I will detail in stages what I did and what happened:

1- Run DDU in safe mode. Uninstall driver and shutdown (for new GPU installation).
2- Replace 880m with 980m.
3- Boot into Windows. Restart with DSE disabled.
4- Install drivers 359.06. BSOD during installation (first time it does this).
This is the exact dump (analyzed in WHOCRASHED):

This was probably caused by the following module: nvlddmkm.sys (0xFFFFF8013020063A)
Bugcheck code: 0x50 (0xFFFFD000217B8000, 0x0, 0xFFFFF8013020063A, 0x0)
Error: PAGE_FAULT_IN_NONPAGED_AREA
Bug check description: This indicates that invalid system memory has been referenced.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys .
Google query: nvlddmkm.sys PAGE_FAULT_IN_NONPAGED_AREA

5- Rolled back on its own and rebooted into Windows. Restart again with DSE disabled.
6- Install driver manually from device manager. Chose the nvcvi.inf file manually. No error. Asks for reboot. BSOD just before Windows login. Same crash dump as above...

I will wait for your suggestions before doing anything else...
I am also attaching the HWINFO screenshot showing 1GB of RAM for the GPU...
http://imgur.com/UyhukZy

 

hmmm... lets use an earlier driver. the 359.06 had so many issues a few months ago, that I was literally telling people the driver is no good.
Can we try a different driver, and then see what happens.

EDIT : wait, let me find a specific driver that I think works good... some are better then others as such, and some older drivers from last summer are just as bad as the 359.xx were.

https://www.dropbox.com/sh/jdacz9ps5iz1m3m/AADpXrkND1g4MuP-OSHFsjI8a?dl=0

this should be a windows 7/8.1 64bit driver, this is modded.

This should be windows 10 version :

https://www.dropbox.com/sh/nrj40rqy7z2db96/AAC7sTRPLhqD5dzHAchKssXya?dl=0

 

361.43 have been reported as broken by many. Anyway I had to mod them as they fail when i try to install them... I think these are modded for dedicated not optimus


Sent from my iPhone using Tapatalk

 

Exactly.

 

Still bsod...


Sent from my iPhone using Tapatalk

 

BSOD... so it gets to loading windows then Bluescreen ? Whats the error message on the BlueScreen ?

 

This is the exact dump (analyzed in WHOCRASHED):

This was probably caused by the following module: nvlddmkm.sys (0xFFFFF8013020063A)
Bugcheck code: 0x50 (0xFFFFD000217B8000, 0x0, 0xFFFFF8013020063A, 0x0)
Error: PAGE_FAULT_IN_NONPAGED_AREA
Bug check description: This indicates that invalid system memory has been referenced.
This appears to be a typical software driver bug and is not likely to be caused by a hardware problem.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys .


Sent from my iPhone using Tapatalk

Man, I think it really is the GPU memory... Take a look at this screenshot:

 

Most of the newer NVIDIA drivers are absolutely horrible. What driver version are you using now?

With 361.43 I even have 0x124 STOP errors pointing to nvlddmkm.sys and I have never in my life seen that. It has always been that a 0x124 related to overclocking a CPU too high with voltage too low, but never relating to the GPU driver. Of course, it also had the page fault in non-paged area error as well, but the 0x124 relating to nvlddmkm.sys is something out of the blue. Don't know what the knuckleheads at NVIDIA are doing to their drivers, but it's not good.

 

I have tried 6 different drivers with the 980m without success... I am back on dedicated mode. No joy. NO software is able to spot the GPU's ram. Speccy only gives me the adapter name (980m) without any other details. As for the advance settings (adapter) in the Resolution options, It shows available memory: 0....
I can't find a single instance online of that happening. I am able to install a driver (granted PC doesn't boot after restarting....) , flash a vbios on the card, even check the GPU temp in HWInfo...
Am I that unlucky?? What the heck could cause the memory modules of a GPU to fail like this?? I don't even overclock!

Woodzstack, can you please PM me what to do regarding warranty?

Thanks!

 

@nick81 - what vBIOS are you using? If you already posted that detail I apologize because I haven't seen it. It's definitely not right that the vRAM is not identifiable, but there may be a firmware/software problem causing it. I have never heard of the problem you are having, but it sounds like (short of a hardware problem) a firmware issue.

 

In my defense of ignorance, I have not tried the latest drivers myself, also did not know if they had any history (already -geez..) of being bad.
Try fox's very old desktop driver for Alienware's, if possible, I would like to verify if older stable 344.xx drivers work.
As for warranty, its not an issue, I will forward you the info you need, but at a time like this when nvidia releases something that 95% of the time is the culprit, going to warranty is just a waste of time, because it might just come back and you wasted time and shipping, just to get instructions to make it work, however, there is a chance, yes, that something is wrong, too much unknown. I will PM you, rest assured everything will be fine.

 

I was on prema 1.1.1 and tried slv's vbios as well tonight. No luck...



Sent from my iPhone using Tapatalk

 

What about the stock vBIOS that came on the GPU? Same thing?

 

I dont remember as I immediately flashed to prema but I think it was an older version of it. Woodzstack might confirm that


Sent from my iPhone using Tapatalk

 

apparently enabling no integrity checks not only allows for installing modded drivers but allows for rootkits to enter the system....so i reverted back.

 

Yeah, I am not worried about that too much. I'm not even using any antivirus software any more.

 

i'd rather be safe and if i have to disable DSE manually when i install a driver then i don't mind.

 

Aha! So that's why ****** Windows Defender was bothering the last few days.

 

my 980m temp runs 45-71 max with thermal pads on..
try reseting your bios and make refresh install of your windows better go windows 8.1 if you gonna use 980m .. then update intel hd in device manager first then restart install nvidia 358 driver after all the restart install the latest nvidia driver from eurocom ithink its 360 something... thats how i got the best fps benchmark i got 3350 score on unigine..

if you cant install driver even you ise dse download woodsztack inf file from his dropbox then paste it on the driver you gonna install ..

trust me im engineer lol...
sorry if i didint fix my messege and check my grammar my phone is lagging hard no time for it

 

yeah you're right im trying to make work the on screen display to use fn f7 but it doesnt work even i do everything like installing the osd first then the 980m .. so i just gave up and reset my bios to get out from the 8 beep then use optimus..

Btw here's my benchmark
no OC
https://www.dropbox.com/s/libtcbi1b2bxwvg/benchmark.png?dl=0

 

I believe it was, indeed.

OSD issue ? What windows are you using ?

 

Anyone knows the size of the screws for NVIDIA X bracket? Mine old screws from AMD won't fit...

 

If you have a Clevo support plate (x-bracket) they use smaller screws that the ones Dell/Alienware uses for both AMD and NVIDIA GPU. I don't know the size measurement, but @woodzstack can likely get them for you or tell you what size they are. If you are not in North America, your closest Clevo reseller can send you a set of them and it is normally very inexpensive.

 

Thx for the reply. the screws' head that came with NVIDIA 980m are really small, they can go pass two holes on the heat sink.

 

Now, that's really odd. The thread size on Clevo heat sink screws is smaller, and so is the screw head. However, I have never seen any Clevo screws with heads so small that they fall through the holes in the heat sink spring retention arms on Alienware heat sinks. Those screws have always had heads large enough to work fine for me.

 

The screws usually depend on the X-bracket attached to the card.

Mr.Fox is right, if you need one I have one, with the screws. Lets say 20$ shipped, via 2-3 day shipping, and you get the screws to match it.

NOW - if its no the X bracket you need, i.e maybe your screws fit the X-bracket you have, but the screws pass THROUGH the heatsink part that attaches to the GPU, then maybe just get some small plastic washers from your hardware store.

 

the dell nvidia x bracket uses the same screws that hold down a few other system components like the bottom cover and the wifi module.

 

Thx a lot! I made some small copper washers so they wont pass through! Really enjoyed your tutorial, I can find everything I need here

 

I try to do my best ! Thanks, Most of this is not even mine, but just things I accumulated thanks to this forum and Tech Inferno too. Thats why i always tell people, if your into laptops, sign up to both !

But your welcome Willoc!