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

All great questions, but no answers. There should be no circumstances... ever... under which the OS or display drivers should even attempt to make changes of any kind to firmware, for any reason whatsoever. This should only be done through the OEM and application should be optional at the sole discretion of the machine owner. Updates or changes at this level are outside the scope of appropriate influence and control of anyone other than the person or entity that owns the hardware, and should require a conscientious and deliberate act to be initiated at the keyboard. It certainly appears that either Micro$haft or NVIDIA have taken a liberty here that was not theirs to take. If one of them did they should be accountable for paying any and all costs to correct the problem for all adversely affected, as well as publicly apologizing and undoing any changes to all systems, globally, that were initiated without express permission, including those that are not visibly affected. Being molested is wrong whether you realize it happened or not.

 

On a Windows 10 sidenote, there's an Unreal Engine 4.0 DX12 Tech Demo out that you can download and run in real time on your GPUs (download link is located on the YouTube link):
http://www.guru3d.com/news-story/unreal-engine-4-elemental-tech-demo-dx12.html

You can run the benchmark in DX11 (if you run the main application icon), or in DX12 if you run the link marked "DX12". It runs OK on my laptop - I got an average of 40fps as recorded by FRAPS using DX11, with minimum fps in the high 20's. Running the DX12 version seemed to be the same fps (FRAPS wouldn't run with DX12), except overall power draw was probably on average 5W less as measured by my Kill-A-Watt - I think this was because I was getting 99% constant GPU utilisation during DX11, and during DX12 I was getting slightly less at 97% constant GPU utilisation, CPU utilisation seemed to be similar between the two versions strangely, although did look that DX12 CPU utilisation only very slightly more evenly distributed between cores. VRAM usage on DX12 was 700MB higher than DX11. Both DX11 & DX12 versions looked the same, so I don't think they're running different settings. All a bit MEH, but happy that the DX12 demo ran well on my laptop regardless!

 

I contacted my local news media this morning, and have submitted my story with a link to this thread, my videos, and the Nvidia forum thread regarding this issue. I refuse to just sit back, and allow Nvidia & Microsoft to ignore us completely. They are destroying multi thousand dollar machines, someone needs to do something about it. If the reporter from this news group doesn't get back to me, I can copy+paste that email to all the other local news groups as well. I'm sure that a like-minded reporter is out there; will keep everyone updated as this unfolds. Hoping for some media coverage, sure beats making time consuming YouTube videos to spread limited word to warn others.

 

Awesome! I alerted the person that @DumbDumb gave us a link for as well. Hopefully, bad press will get Micro$haft and NVIDIA more interested in fixing their mistakes. I feel bad for myself, all of the others affected, and even for Dell/Alienware... they are victims as much as we are and may be picking up the tab for in-warranty machines that have been messed up by the incompetence of third-party "partners" that caused it.

 

I hope M$ will cover those LCD panels for you

 

I think that I/we would be holding our breath to the point of asphyxiation if we were to depend on Micro$oft or NVIDIA to take responsibility for this. If history holds true, it will probably never happen.

But, thanks to the overwhelming generosity of members of this community, I have replacement panels for the M18xR1 and M18xR2 on their way to my door already. I am not going to name names, but they know who they are and I am very touched by the kindness. I may have a replacement LCD for the Alienware 18 also, just waiting for confirmation that one is available.

The mot important thing is that we identify why this is happening and put a stop to it. I feel very bad for people that may be experiencing this that have not found his thread, or the one in GeForce Forums, don't know why it is happening and don't have a warranty. If I had only one affected Alienware and did not have three in a row that were affected, I would have even assumed (and actually did when the first one died) that it was a random hardware failure.

 

I just want to say " thank-you" to those generous souls who helped you out brother Fox. You are absolutely awesome and so are they.
We stay strong when we remain united; and this community is priceless; offering some of the most valuable insight on tech problem solving.

 

Cool benchmark. Looks great, too. Shame the engine does not support SLI. Ran it on the Panther without SLI using 980M (Windows 7) and got this...

Code:

2015-08-07 14:21:05 - UE4Game
Frames: 13946 - Time: 168013ms - Avg: 83.005 - Min: 33 - Max: 264

Borrowing the one for DX12, I created a shortcut and added -fullscreen and that fixed the issue of it running in window mode (hate vsync and window mode, LOL).

Code:

%windir%\system32\cmd.exe /q/c start ElementalDemo.exe -fullscreen -novsync

No plans at the present to entertain Windoze OS X on the Panther. Waiting for Eurocom to test Windoze OS X first and find out if anything gets screwed up instead of being a guinea pig again.

 

Just to give you another data point... I am using Windows 10 with nvidia driver version 353.62 on an Alienware 17 with 880m and 120 Hz 3D screen.
I've had no problems whatsoever. Everything still working fine.

 

Mr. Fox, looks like its true that EDID Firmware was scewed. That guy reflashed EDID firmware and his laptop is working again. He claims that half of firmware was blank so Win10 or Nvidia drivers erased it.

I asked him about more info but it's like it's a pain for him to explain everything in a normal manner.

 

That's good. If it killed every Alienware in existence that would be truly catastrophic. Thanks for sharing. It's always good to have as many data points as possible.

Since you have eDP (versus LVDS) and probably running in UEFI you may never experience the disaster some of us have and others may soon be cursed with. It doesn't make whatever Micro$haft and/or NVIDIA have done any less sinister, but it's awesome that their incompetence has not hurt you.

Thank you, Brother @James D for posting the update. I was pretty certain that was happening, but had nothing but a hunch to base it on. I'm wondering what he used to dump the firmware from another display and how he flashed it to the damaged one? It would be awesome to be able to fix the problem, but I don't have a clue where or how to start.

I also wonder what, if anything, can be done to stop the problem from coming back again other than avoiding Windows 10 and/or certain NVIDIA drivers. If we knew what part of Windows 10, which Windows Update, or what part of the NVIDIA drivers was doing the damage, we might be able to block it.

 

That's least I could do. Additional info:
Panel: lp173fw1 (tl) (b3)
Firmware: Direct link on N173HGE-L11_C1 but possibly you can use any with same diagonal and resolution etc (he did that). Also you can check compatibility model list. Firmware source
How-To: guy used programmator TL 866. It is possible to flash through LVDS cable and trough chip directly. Guy sent LCD to SC where they soldered 4 cords to 4 legs of memory chip which he connected to programator. For LVDS you need to find LVDS 40pin, ring 4-th, 6-th, 7-th and 10-th contact and put it into programator. Video example: https://www.youtube.com/watch?v=cNsHQhSXtz8

Believe or not but he claims he is still using Windows 10 now. I guess that's because he knows how to fix it now.

 

oddly enough i don't see any release notes on the driver page....where's it at?

 

Hello all, I just want so share some info. I'm running a Clevo P377SMA machine, 970M SLI. I've been using Windows 10 + 353.62 drivers since July's 29th and so far it's been great. I don't know why but I have a hunch this catastrophe may be a problem related to Microsoft forced updates (via Windows Update). When I learned of this, first thing I did when I updated to Windows 10 was using DDU to disable forced graphic drivers updates and made a clean install of the 353.62 drivers before any update could mess up my drivers. Also, I always set my brightness to 85%, just in case... So far it's great, but I know I'm running a different model and all, so this all could be jazz.

 

With both GPUs removed I get the 8 beeps, as per first post. I will try removing the RAM and see what happens.

EDIT: All RAM removed - 2 beeps. A different stick one at a time in the first slot - 8 beeps.

 

Thought ' Interesting thread for detectiving' ... 37 pages later ...:

These are the three different AW part nrs. posted so far. Mind that these are 7-character shorthand, not actual panels. The part nr. would be after "Data string" (from my panel):



This is from a Clevo and correctly lists "B173HW01 V5" (matte version of the P570WM's glossy V4). Dell, however, in its infinite wisdom, does not comply to standards (anymore);



That string is not a panel, but a Dell inventory number for a range of 'compatible' panels (easier replacement/ordering etc.). Searching Samsung 18.4" yields 22 panels, which have either 184HL, 184KT or 184HT as part of their full string. Checking the inventory number leads us back; there is only one HL and, though there are multiple HTs, there's only a single RGB-led. Furthermore, these two are the only LED types, the rest are CCFL (also more robust). Together with the LG, this summarises the three panels so far:

1. V @60Hz LTM184HL01
2. X @64Hz LTN184HT02-S01
3. X @??Hz LG LP173WF1-TLB3

Now, going to guess here that the LG also ran beyond 60Hz spec., due to the corrupted EDID (usually stored in EEPROM, but not necessarily so). Both questions (refresh rate and EDID) can be answered:

• Refresh rate; open Blur Busters' webpage (real-time refresh rate checker).
• EDID: Running MonInfo will tell you right away whether it's the EEPROM:

Spoiler: LTN184HT02-S01 of user [Nikos]

This is proper and pulled from registry, which in turn is pulled from the EEPROM upon boot. If corruption is the cause, then there'll be discrepancies/nonsensical data here. Now, @Daniel1983 can still access the 184HT while running; will be interesting to see that screenshot.

Also; HMDI-connect will force 60Hz, including that of the internal display, bypassing any corrupted values. At least, those concerning refresh rate ... Running crazy values, unfortunately, might also have damaged Mr. Fox's panels, considering connecting HDMI device doesn't work for him.

That the failed panel runs brighter does not, counter intuitively, have to do with actual brightness control; most panel use pwm control and here 100% = 100%, never more. Yet, increased refresh rate has the same effect; less time to dim before the next flash. Perhaps @Solo wing can check Blur Busters' page; see if it goes over 120Hz. That, too, would increase brightness.

Even if Mr. Fox's and Daniel1983's panels have a corrupted eeprom, there's still a potentially easy fix for the time being; overwrite registry with the proper values ( @[Nikos] can export those) and set the keys to 'read only' right after that, without rebooting. Any reboot and the faulty eeprom values will be written again, same with an OS reinstall. Think you may need safeboot for this, or a WindowsPE rescue disk, anything that will let you access and edit the registry while not in use. Registry values are stored under:

Code:

[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\DISPLAY\AUO159D]

Spoiler: B173HW01 V5 (EDID)

Where, of course, AUO159D -> SEC5448. Apart from that, there's also the driver control key:

Code:

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers\Configuration

Spoiler: B173HW01 V5 (driver)

Would interesting to see what Daniel1983's system says here for VSyncFreq; perhaps 60 with HDMI and something else without. There's even a possibility it's not the edid after all, but just those driver values (fixed in a jiffy, if true ...). Can auto-export (and import -> 'regedit /s monitor.reg') registry even when sailing blind. Place AUTOEXEC.BAT in root with something like this:

Code:

regedit /e file1.reg "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\DISPLAY\"
regedit /e file2.reg "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers\Configuration\"

This may actually be necessary as there's a subkey 'Properties' under 'DISPLAY' section which cannot be written or exported while active ("Access denied"). Could take ownership, but sometimes this has to be set to 'System', so ... make backup first (and prepare autoexec.bat write-back).

This is obviously a temporary solution; proper fix would be rewritten edid. Have to take care here; RGB led is pretty rare and not sure whether it'll play kind to a WLED edid. Fortunately, @[Nikos] (again ) can pull the correct one with MonInfo ( highlight proper panel -> File -> Save as -> dspinfo.bin).

Ps.
There's nothing wrong with overclocking a display in itself. It just has to done with correct timings and the proper channels; my Dell D800's WUXGA runs 185Hz ( very old ancient Sharp TN).

 

here's a dell AUO panel



EDIT: am i right in saying that dell has been rewriting the EEPROM of panels prior to building and shipping of laptops? if so then why hasn't screen burn out happened earlier? why windows 10 + 353.62 drivers in particular? can panels that dell has flashed be flashed to stock firmware or as the panels that dell get are using proprietory/custom firmware?

 

Glad to hear about the panels being replaced. Also glad to hear about the issue slowly being tracked down as well.

 

@t456

 

Hmm... that sure doesn't seem right @Daniel1983. Microsoft does not make LVDS or eDP monitors. The specs are also screwed up.

So, it looks to me like Windows 10 is, in fact, responsible for the damage. Might be a rogue update as suggested.

@t456 - great info you posted. I cannot boot my systems, possible because they are set to Legacy BIOS. When I get the replacement panels, I will boot them up with the good LCD and change the BIOS to UEFI without Legacy support and reconnect the messed up panels and see if I can boot from HDMI or mini-DP.

 

My thoughts exactly. I feel like I should be making a live video of this because it is hard to believe! Can this be fixed; if so how?

 

someone needs to dumb it down a bit LOL

 

Are you running Windows 10

 

Yes, up til I got the 8 beep issue.

 

Some thoughts.. When will AntiVirus companies start incorporating protection from corrupt drivers and malicious code being written to LCD's.?. Last I checked my BitDefender subscription doesn't offer that. Once upon a time I believed being behind an AlphaShield hardware firewall I was invincible. Times have changed.

A quote comes to mind: "Evil can only prevail when good men do nothing"


Sent from my iPhone using Tapatalk

 

To have one laptop break down sucks big time.....but three Alienwares..... I feel terrible for you...

I recently purchased an M18x R1 and after having upgraded to Windows 10 there was one particular bad day during testing that the AMD drivers had to be continually recovered with the screen flickering badly.
I didn't think much of it at the time and once I experienced a blue screen things returned to normal.

It might just be a coincidence but I thought it worth a mention as I'm beginning to believe it's not solely down to Nvidia cards.

EDIT : Using AMD 15.7 drivers

 

Join the 8 Beeps club.

http://forum.notebookreview.com/threads/windows-10-upgrade-warning-for-alienware-owners.779449/

 

This user is having the same issue

http://forum.notebookreview.com/threads/m18x-r2-8-beeps.779728/

 

Good performance you got there from a single 980M - double the performance of my GPU, but to be expected! I modified the shortcut targets to my own ends too.

 

The more I see the more I am starting to believe that Micro$haft's Windows 10 abortion is the root of our evil circumstances. I just hope the cause can be isolated so the responsible entity can be held accountable for their underhandedness, or incompetence, whichever the case may be.

 

Nope. No fix at this point. Buy a new LCD and go back to Windows 7 or 8 is the only option at this point. If you replace the LCD and keep Windows 10 it will end up screwed up just like the one you have now.

This is happening to Alienware 18 owners as well. At least one Alienware 17 owner has also been affected.

Brother @Arestavo - please also post in GeForce Forums to document your victimization. Here is a link: Windows 10 + Official 353.62 Drivers are Killing Samsung and LG Notebook LCD Display Panels | GeForce Forum

 

Ok, so it looks like my post on the M18X R2 forum wound up here. I'll expand a little more now that I see what is going on.

Specs in signature (M18X R2). I installed Win 10 with the iGPU because it would fail with the 780M's enabled. Switched over to the 780M's and installed the modded 353.62 drivers from laptopdrivers2go. Updated and downloaded some games, played about 30 minutes of GTA V - all told maybe 5 hours of use. Shut off the PC, and 2 days later tried to boot it up again to 8 beeps, even with the 780M's removed.

Gott verdammt it so much.

EDIT: Which company, Nvidia or Microsoft, do I send my ~150 USD bill to for a new LCD panel?

 

Good question. There's a spot for you to express your displeasure over here also [ LINK ]

 

This is a long long shot, but I may as well communicate it. I was having problems cold booting with Windows 10, and quite often getting 5 beeps (GPU failure, yep and it's not the 8 beeps you're getting) amoungst other startup issues, but I turned off the "Fast Start-Up" option in the Power Options and I don't have this problem anymore - see attached pic for where to find it. Maybe Fast Start-Up is writing something to the BIOS or the way the laptop starts up or something.

 

@Mr. Fox , I think you should see this.
Apologies for taking off-topic, I couldn't find an active Win7 thread.

The Draconian Mafia empire strikes back (BGA was first), a huuge blow to all Win7 users with the Intel 100 series / Skylake machines.

Spoiler: Intel / Microsoft retire Win7

Source : WCCF tech

Spoiler: Anandtech's take on Skylake + Win7

Source : here

Edit : A doubt, can this fix (win7 usb3.0 only installation) be applied on the skylake machines ?

 

The install is crippled not nuked, big difference.

 

Edited my post, thanks.

 

Posted in that thread Mr. Fox. I also submitted a ticket to Nvidia - everyone with this problem should get it documented with the Nvidia support team. It will be useful if Nvidia refuses to acknowledge this problem and compensate (or fix the LCD) those affected, and a class action lawsuit is required.

EDIT: I will try putting everything back together (with 780M's) and hook up my HDMI TV to the laptop to see if that boots.

EDIT 2: I don't seem to get HMDI out (which could just be my TV as it is slow to see inputs) with either the on-board LCD hooked up (8 beeps) or disconnected, but I DO GET A SPLASH SCREEN (SLV7s VBIOS splash) with a VGA monitor and the on-board LCD hooked up - but I STILL get that looping 8 beeps.

EDIT 3: Mr. Fox, any chance that anyone (such as SLV7) could work up a flash recovery fix if this turns out to be EEPROM corruption for the LCD?

 

@Daniel1983 can you also show EDID info of other 2 monitors?

 

Possible, but not likely (more effort). Eeprom is optional, not mandatory; settings can be set via bios too. It's not good for end-users, of course; values in bios may be incompatible with replacement panel, crippling update routes. Yours is any one these eight, only way to verify which is the sticker on the back. If it's a V4 and you want to sell it (after your 120Hz upgrade); send me a pm . Can also rule out eeprom vs. bios by running it on a Clevo. Or, perhaps, hex-search the AW bios for the corresponding string:

Maybe because 10 forces Secure Boot by default; edid eeprom (and vbios) are a separate 'load'. If 10 blocks that (because it's considered a rootkit) then all that's left are the settings from the Dell bios (booting legacy 7 or xp should fix it then). Either that, or, indeed, these specific drivers send a write command, screwing up the edid parameters (trouble follows panel). All very speculative as of yet. Better see what that Samsung (or LG) edid says.

Thanks, but you want the third listing ("SDC4C48"), this is an external monitor. Maybe check the "SAM0469" too. There's also the possibility it's only the nvidia driver that has the wrong values. Can you run the registry export and copy/paste the contents of the two .txt files?

Code:

regedit /e edid.txt "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\DISPLAY\"
regedit /e driver.txt "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers\"

Or just run the .bat: export edid and driver values. Don't need to use MonInfo then; it's the same thing.

The 60Hz is as expected and you cannot verify overclock anymore, either; can only be checked without hdmi hooked up. However, Solo wing can check refresh rate, except it'll be for a 120Hz panel.

Do you have a programmer kit + clip? Eeprom shouldn't be difficult to locate and you could pull it off the new panel and write back to the failed ones. Would also be interesting to see what 'edid-decode' gives; it's straight from eeprom instead of indirectly via registry. Hopefully the damage wasn't permanent (expensive things, these 18.4"s ).

 

@Mr Fox
Is there anywhere we can post to Microsoft or Nvidia about this?
My m18x r1 just bricked last night and I stumbled across this thread. I wish I had waited to upgrade now. 20/20 hindsight.

 

Sorry it happened to you also. Please post in the thread I started at GeForce Forum.

Here is a link: Windows 10 + Official 353.62 Drivers are Killing Samsung and LG Notebook LCD Display Panels | GeForce Forum

I messaged Gabe Aul on Twitter twice and sent him an email with links to the above thread. He has not responded, and might not. But, I'm reasonably certain Micro$oft is aware of it.

 

Already done
I saw your link when I was reading through hoping for a fix.

 

Everyone with this issue - Call up Alienware @ 1-800-254-3692 and open a ticket with them even if you are out of warranty. Use the tech support option 4. If you get a rep that seems to not care and won't open up a ticket number, call back - I had to once to get a rep that cared.

This will help establish trend lines and get Dell/Alienware to lean on Microsoft/Nvidia. Which will hopefully get the culprit to foot our LCD replacement costs.

 

What do we say? That we installed windows 10 and now our displays are fried.
I'm not good at the whole support thing forgive my ignorance.

 

Excellent advice. I agree with the recommendation.

If anyone that calls in can find out if there is a master ticket or case number for tracking, posting it here may help other not have to call back and talk to multiple reps. If there is one I will update the opening post here and at GeForce Forum.

Yes. Point them here and the thread at GeForce Forum and let them know you want them to track it. If there is a master ticket or case number for tracking and we can get that, posting it here may help other not have to call back and talk to multiple reps. If there is one I will update the opening post here and at GeForce Forum.

This thread and the other one have a high view count for only 5 days, so I suspect Micro$oft, NVIDIA, Dell/Alienware and perhaps other major hitters in the industry know about this. Hopefully, the media will pick it up and start blabbering about all over the internet soon. Probably would not hurt to post links to this thread and the GeForce Forum on Facebook and tweet links on twitter as well.

Spoiler: Hyperlinks for Facebook and Twitter

https://forums.geforce.com/default/...g-samsung-and-lg-notebook-lcd-display-panels/

http://forum.notebookreview.com/threads/windows-10-upgrade-warning-for-alienware-owners.779449/

 

It's my own personal ticket number for this issue, but please feel free to use it as a reference: 914989593

@Sspawn26 - Call up and let them know that you aren't looking for free warranty work (if you are out of warranty), but you want to document this issue. Have your express service number and service tag ready to give to them. You will want to specify the operating system (Windows 10), your graphics card vendor (Nvidia), the Nvidia driver version (353.62), and the post code error (8 beeps).

 

[QUOTE="Thanks, but you want the third listing ("SDC4C48"), this is an external monitor. Maybe check the "SAM0469" too. There's also the possibility it's only the nvidia driver that has the wrong values. Can you run the registry export and copy/paste the contents of the two .txt files?

Code:

regedit /e edid.txt "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Enum\DISPLAY\"
regedit /e driver.txt "HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\GraphicsDrivers\"

Or just run the .bat: export edid and driver values. Don't need to use MonInfo then; it's the same thing.

The 60Hz is as expected and you cannot verify overclock anymore, either; can only be checked without hdmi hooked up. However, Solo wing can check refresh rate, except it'll be for a 120Hz panel.
Do you have a programmer kit + clip? Eeprom shouldn't be difficult to locate and you could pull it off the new panel and write back to the failed ones. Would also be interesting to see what 'edid-decode' gives; it's straight from eeprom instead of indirectly via registry. Hopefully the damage wasn't permanent (expensive things, these 18.4"s ).[/QUOTE]




Here is what the above looks like on my other AW18 that is NOT upgraded to Win 10:

 

Yeah, your registry information looks valid which means that if something is wrong then it is definitely in the panel itself, hense why broken laptop doesn't have Real-Time monitor EDID data, only Registry one.
You need a programmator. The valid Firmware you already have. You need just to flash it on a chip.

 

I don't know how to do that fix. If someone could provide clear detailed instructions I wouldn't mind attempting it to see if it works.