*** Windows 10 + NVIDIA WHQL Drivers are Killing Alienware and Clevo LCD Panels ***

Straight from their "knowledge Base"
https://www.dell.com/support/article/us/en/19/SLN297954/EN

 

Omg @Mr.Fox, didn't know the situation became so bad for you... After reading this thread, I have a feeling that Windows 10 with 353.62 might have killed my 970M.. Damm I think I am going to stick with 8.1..

Also anyone knows how to claim compensation from Microsoft for this WHQL crap?

Sent from my LG-H811 using Tapatalk

 

.

 

@Mr. Fox separate thread -> https://forums.geforce.com/default/board/33/geforce-drivers/ thread title ... include Samsung LCD screens

-> https://twitter.com/gabeaul <-

 

here is an older, but similar, issue.

http://forum.notebookreview.com/thr...light-syncing-issue-cause-a-blown-lcd.627287/

 

WAIT???!! If I'm looking at this correctly my AW18 is not supported?!?!! It mentions AW18 R1 is that the older (1st model) that is at least 4 years old? I've seen where my PC has been called the R3. This R stuff always confuses me.

 

AW18 is what you have, it's the AW18 R1 aka M18X R3

 

No. Per that website, the AW 18 is supported. Not the AW m18x R1 & R2.

It's part of their naming scheme to add R1 even if there is no R2. But I could be wrong and they are in fact going to release an AW 18 R2 with skylake.... BGA all the way round.

 

Alienware 18 R1 is the A18 - the latest one. It's been M18xR1 -> M18xR2 -> A18.

When it refers to these, the following are the newer laptops, 2013 onwards it seems (none of the M18x nor M17x nor M15x series are in the list of supported):
Model
Alienware 13
Alienware 14
Alienware 15
Alienware 17 R1
Alienware 17 R2
Alienware 18 R1

 

I'm not sure which decision makes me want to hurl more.

Alienware ceasing production of the 18 inch model

Or a fully soldered BGA Alienware 18 inch model



......................................They are both very distasteful options........................................

 

I don't know, looks like these laptops or maybe just the panels have a design flaw, look the next thread here on this forum, it's from 2013, not nvidia driver or even more win10 related:
M17xR3 8 Beeps of Death

And here is his last post about his problem:

It's very suspicious, just a couple things to discard.
Maybe is a combination of some other bios feature (like the "non-turbo flex support" case), the OS that manages that feature in a different way and the driver:
-There is a similar case (same NV driver version) but on W7?
-There is other people have the same laptops, same OS and same driver but not suffer the problem?

On both scenarios is a good idea to compare bios version and configuration, if both cases happen then people must avoid 353.62 and newer drivers and report this to Dell and Nvidia, they must investigate what's wrong here.

 

^^^Nope, sorry... none of those things apply here. Windows 10 + NVIDIA drivers killed 3 Alienware laptops screens (all are Samsung displays attached to NVIDIA with 353.62 drivers) within 24 hours or less of use. ^^^

Yeah, I examined them closely and I could see no sign of physical damage.

This is completely irrelevant. What they recommend is for their own good and they say the same thing whenever a new OS is released. It's so they can avoid having to support the new OS on old hardware. Plus, one of the machines that is supported has a Samsung display that died from the same cause as the two displays on the other machine (which is basically the same display) that is not on the list. And, for the record... I don't have a warranty and even if I did this would NOT be Alienware's fault. Their product is just fine... Windows 10 + 353.62 drivers obviously are not OK for these display panels. I don't know why, but it is what it is.

My GPUs are perfectly fine. It is only the 3 display panels that have failed. Good luck getting Micro$haft to do anything.

I think I would go one step further and say NOT AN OPTION... PERIOD! Remember, can't fix stupid... that's terminal. Releasing a BGA 18 would be visible proof that the brain cancer has metastasized.

 

Don't the MSI GT80 and the Dell XPS 18 use the same display panel as the 18" Alienwares? I know the XPS 18 does, but not extremely sure the GT80 does. I'm trying to search around to see if those two computers are having the same issues. Maybe not the XPS 18 since it uses an iGPU (you tested with an iGPU and it seemed fine) but maybe with the GT80. Worth trying to see if the driver or OS doesn't play nice with the panel among other computers.

 

I'm not sure, but that is entirely possible. Yes, everything was fine on Intel Graphics. When I switched the 18 over to GTX 780M SLI and installed NVIDIA's latest 353.62 drivers is when it died.

 

I looked at as going to be a problem for their warranty dept when people who do have those panels and drivers update to Win10. Even though they say its okay to update. Maybe this could get Alienware mad at Microsoft for you?

 

Damn, its sad to see that your LCD got broken 3 times, esp. your Alienware 18. Guess I go stick with Windows 7 on my AW14.
Also any news from AW representative?

 

Well, I did find something for the GT80 from MSI's forum.

It probably is the display after all, or the display has a higher chance of failure compared to other displays Alienware (and MSI) use in their smaller laptops. And that's not good.

 

oh cool, i am assuming that your 880M is stock/compatible? that's probably why it isn't downloading for my 980M cos it isn't stock...that might be beneficial

 

Noticed insane screen tear and corruption cold booting on W10 on my 120Hz screen ... She is alive still ... Downgrading to 353.30 and will check if the same issue will rise up

Scared like heck now

 

if it can last 3-7 days without lcd failure then i think it's the nvidia 353.62 driver easy.

 

I hope it is as simple as a driver downgrade

 

i hope so too cos i want to use the 353.38 hotfix driver cos that's hasn't caused any issues and it fixes the TDR issue with chrome.

 

Try turning off the Fast Start-Up option in the Power Options menu - the option is hidden in there somewhere. I was having even more extreme cold boot issues (although not the screen corruption you talk of) until I turned that off (I still have a strange Cold Boot Low GPU Performance problem though that is solved on a restart, but at least it boots now!).

EDIT: here's the link to where you can disable Fast Start-Up: Control Panel\Hardware and Sound\Power Options\System Settings

 

Old driver installed I hope im safe now ... Went in with the updates hiding tool and I hid the Nvidia update and all other driver updates the tool found as I am really not ready to lose my currently only active machine to developers incompetence

I couldn't find that option anywhere ... Maybe cuz I have fast boot disabled at Bios level ?

 

I sure would like to know what the "variety of problems" are, if they are Windows 10 display related, but I am not spotting anything on their forums or in Google searches. Maybe they are deleting posts and talking privately to those who are affected.

 

I think the BIOS fast boot option is something different to the Fast Start-Up option in the OS. Here's a pic showing you where you can find it, you make sure that the little tick box at the bottom of the screen is not checked for Fast Start-Up: (you have to click "Change Settings that are Currently Unavailable" at the top of the screen before it will let you check or uncheck the boxes at the bottom):

 

Yeah, I had no luck finding any specific problems. Not here, MSI forums, or Nvidia's forum. That said, at least their consumers listen. We, on the other hand, experiment and, well, yeah...
*Crosses fingers, prays, and hopes for the best

 

Oh I see ... Disabled now

 

Excellent! Let me know if it solves the cold boot screen corruption you were getting - it's just a punt from me because I solved my slightly different cold boot issues by turning off that option.

 

The driver downgrade solved the screen corruption ... but wow disabling the fast boot made the laptop boot much faster from a shutdown ... Good job M$ ... /facepalm

 

Ahem... I think you quoted the wrong person.

 

Good lord I see that now ... Dealing with dumb OSs made me dumber ir seems :'(

 

Guys, how does this look? https://forums.geforce.com/default/...rivers-has-killed-three-samsung-lcd-displays/

If you post a reply, please don't make comments that portray this as an Alienware problem, because it is not. There is clearly something about Windows 10 and/or 353.62 drivers that is not a problem Dell/Alienware created. If any hardware device is to blame, I think it has to be the Samsung display panels. I suspect more than anything it is an NVIDIA driver problem. Remember, this comes after 6 months of 780M SLI throttling problems using drivers released after February. That issue may not be related to the display panels being killed, but there is clearly an ongoing issue with NVIDIA drivers not functioning properly.

Thanks for the tip.

 

Nicely written ... Lets see how they react "if they do"

 

Yeah, it's stock.

Will bump the thread when it drops.

 

i wonder, is it just the samsung screens? what brand of screen that failed in the AW17?

 

I'm not sure. Maybe @scarastisis will find out for us and post in this thread.

 

mine is a stock AUO panel.

 

My Clevo P570WM has a AUO panel as well. I'm not willing to risk destroying it until I see a few examples of success. I may reach out to Eurocom and find out if they have validated Windows 10 on this model yet, with the 60Hz panel.

Let me get the display model info and see if it is the same as your Ranger 17.

Edit: You can use this to check... Monitor Asset Manager | EnTech Taiwan

Monitor
Manufacturer............. AUO
Plug and Play ID......... AUO149D
Data string.............. B173HW01 V4

 

my AUO panel model is B173HW02 V.1

we still haven't heard from 120hz peeps.

EDIT: all we know is that it breaks samsung 60hz panels.

 

So after drooling on myself in shock at this disaster for about an hour I have a couple of things in my head. First of all, I thought this was just an issue with this driver if you are running Windows 10 but Mr. Fox, did you say one of your machines was not running W10? If that's the case, this driver may not be safe for many folks.

My other thought. To everyone this has happened to, are all of you benchmarkers with a current overclock setup? Just curious. I know they supposedly enabled overclocking again but I don't believe for 1 minute that they actually want people to be able to overclock. Just wondering if the issues with drivers are related to them noticing overclock or altered settings and then crippling or not allowing things to operate correctly while those altered settings are in place. This would of course apply to all the recent driver issues, not just this one.

 

Looks like you may have a newer revision of the same display used in my Clevo. I updated my previous post with a link to download a utility that will scan the info...
Monitor
Manufacturer............. AUO
Plug and Play ID......... AUO149D
Data string.............. B173HW01 V4

See the sequence of events in the opening post for clarification. The only LCD failures occurred on Windows 10 with 353.62 drivers. One of the machines with a dead display (M18xR1) was a LCD parts donor after the first failure. No overclocking was involved. The display panels were running at stock specifications, 60Hz and 1920x1080, (unless the drivers were causing them to run at a higher frequency without my knowledge). The drivers I was using do not support overclocking with 780M SLI. They throttle severely above stock GPU clock speeds, but even if that were not so that should not cause the LCD to fail.

 

I was home yesterday afternoon and I won't be home till later today, but as soon as I can I'll take off the bezel and let you know what model the panel is


Sent from my iPhone using Tapatalk

 

mine was manufactured is 2013 yet it still doesn't prove that AUO panels are not affected by these drivers. we need more info from other cases like the AW17 and the lenovo Y50 and see what panels are being used there.

it'll give us an idea though it certainly won't be definitive.

 

@svl7's Alienware 17 R1 120Hz Samsung SEC5044, like the M17xR4 (LightBoost).

 

if it turns out that only specific panels are affected we shouldn't have to cherry pick the panels to make sure they work and this is clearly the fault of the driver supplier: nvidia.

do you know the exact model #?

EDIT: if you have an affected alienware M18XR1, M18XR2, AW18 R1, AW17 R1 please specify your panel's model number and manufacturer.

 

This is seriously bad. Can't believe they wouldn't test something like this ahead of time.

 

so if these drivers have caused your laptop to fail (or not if you're lucky) then post your panel's model # and manufacturer.

 

Well, these machines are all end-of-life models. Alienware says they have tested the 18, so they probably have. If the problem is specific to 353.62 drivers (which it might be) and that driver was not released at the time of their testing, that could explain the situation. NVIDIA uses the general public for testing, so the chips just fall where they may. There is no way they could test every conceivable hardware configuration even if they want to. But, yes... it is very bad.