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

I'm pretty sure he turns off updates during install, then only installs specific updates that he chooses and downloads from Microsoft's website (not update manager).

 

I'm pretty sure I remember him saying something like that but I'm just trying to confirm my hypothesis about those Windows 10 ready updates. Everything feels like that's the corporate on Windows 7 but I want to be completely sure before I go and repair a screen. I don't have the cash to sink into a laptop that keeps breaking.

I'm definitely not going to run Precision X but I wonder if we could still keep it fully updated for securities sake. I looked into Steam OS a little more and I just don't think it's ready yet, maybe never will.


Sent from my iPhone using Tapatalk

 

Yes, @Arestavo is correct. I have been disabling Automatic Windows Updates for a couple of years on all of my systems because I got sick of having to reinstall Windows due to updates causing weird issues and lowering my overclocked benchmark scores. I used to have some of the most important security updates installed (mirrored only what was deemed important enough for my employer's enterprise computer fleet) before all of these Windows 10 LCD bricking problems began, but I am now only running Windows 7 SP1 and Windows 8.1 Update 1 with no additional updates installed.

I wouldn't lose any sleep over not having the plethora of Windows Updates installed. I have had zero problems running with only critical security updates installed. You can manually check and download those. Everything else is mostly garbage that is not needed. Automatic Windows Updates install indiscriminately and the optional and important updates won't do anything to make your system secure. If you stay away from questionable web sites and don't download pirated software and tons of torrents, you have a low risk. If you keep a nice, clean Windows drive image, you can just restore that and kill the malware bugs if you get hit.

 

No; the '-5' and '-6' are only there to override device id mismatch (existing vbios vs. vbios to-flash). There's nothing the nvflash can do to wipe non-vbios eeprom area, except for using a padded vbios.

Anyway, perhaps there's need to send me an MXM card (or desolder and use programmer yourself). Noticed this nvflash switch:

Code:

NVIDIA Firmware Update Utility (Version 5.163)
-- Commands and Options --

display      d [bytes]          Display 256 the first bytes of the EEPROM (default is 256 bytes)

Neat ... let's misappropriate that feature . Already know which vbios eeprom my system uses (from Clevo's Service Manual), but that's a BGA, so ... let's find out the size of the eeprom using nvflash itself:

Code:

nvflash --pciblocks

Identifying EEPROM...
EEPROM ID (C8,4012) : GD GD25Q20 2.7-3.6V 2048Kx1S, page
Reading adapter firmware image...
Location        Size Type        
000000:0195FF 103936 x86            

Now, the '2048K' means 2,048 Kbit (= 2 MBit) and -d switch expects bytes. Thus; (2048 x 1024)/8 = 262,144 bytes. The '--pciblocks' tells us the vbios currently stored is only 103,936 bytes, which leaves 60.4% empty non-vbios/non-nvram eeprom space ... (remember that nvram is part-and-parcel of the vbios, not something that resides outside of it).

Let's proceed to read all 262,144 bytes and store these to a text file:

Code:

nvflash -d 262144 > eeprom.txt

The eeprom.txt is not hex (like a rom/bin), but ascii type. Could just check it 'as is', but why not strip fluff and convert it to a proper 2 MBit rom:

Spoiler: nvflash - extract full eeprom

Checked the remainder of my vbios eeprom; all blank 'FF', so no hidden code here. However, that does entail fate in nvidia's nvflash tool; it might simply read the vbios only and pad the rest with blanks ... we'll see how it goes with the MXMs of affected systems. Can always de-solder and use programmer after all, which is the only 100% accurate method.

As a side-note, nvflash also has an eeprom write-protect option ... :

Code:

nvflash --protecton

Identifying EEPROM...
EEPROM ID (C8,4012) : GD GD25Q20 2.7-3.6V 2048Kx1S, page
Setting EEPROM software protect setting...
Write protecting EEPROM complete.

... and to turn that off again:

Code:

nvflash --protectoff

Identifying EEPROM...
EEPROM ID (C8,4012) : GD GD25Q20 2.7-3.6V 2048Kx1S, page
Setting EEPROM software protect setting...
Remove EEPROM write protect complete.

Or use '-w' and '-r', respectively.

 

thanks for clarifying!

 

Sspawn26, exactly what I was thinking

+rep

 

Hi all,

First of all thanks for everybody for the wealth of knowledge in this thread, specially Mr Fox's and t456's work has been extremely helpful. I've spent the past 4 hours trying to read maybe 10% of what you've discussed but some things are still not clear.

I have a m18xR1 with 580m sli. I've upgraded to win10 mid August and have been experiencing random black screen reboots since (with random intervals). My display did not get bricked - at this point lol. I tried everything and finally landed on trying to increase the voltage as referenced by multiple forums. I tried both EVGA Precisoin X and MSI afterburner. Funnily, was not able to adjust my voltage, lol, but was able to brick my display last night , yay I get the 8 beep and can't boot with external screen.

What's not clear to me is what are the steps and parts that I need to get to the point when I can boot Linux to fix my EDID.

Your help is much appreciated.

 

First off, you need to get a hold of a new screen via buying or borrowing. Then you need to put Linux onto at bootable USB drive. Install the new screen in such a way that everything is still plugged in, but you can easily hot swap the screen when you need to.

Once you've installed your new screen, you will be able to boot again. But please make sure not to boot into windows again, because you run the risk of breaking your screen again due to EDID corruption. Instead choose "boot from USB" in the Bios and install Linux from that handy bootable Linux drive you made.

The rest of the information is found on page 150 by @t456 and a bit more simplified by @Mr. Fox on page 158. Just follow their instructions and you should be fine.

If you had to buy a new screen, at least you can sell it after your original is repaired.

Also note that if you install Windows again, don't go back to 10. Use 8.1 or 7 instead. Also don't let either update to the point of "Windows 10 ready" or use Precision X. Myself, and a couple others replaced our screens and still had it brick after fresh installs because we didn't know not to do that at the time.


Sent from my iPhone using Tapatalk

 

Thank you!

So there's no way around getting a second screen it sounds like

I'm a little nervous about the screen change but I'll look up a few videos on it.

Thanks again!

P.s.: the beauty of is that I'll be still back to my random black screen reboots once I fix my display been trying to fix that for months now

 

Are you sure it's not a Windows 10 thing? My old ladies screen freezes black for like 45 or more seconds before it finally boots. It didn't do that on 7 or 8.1. As soon as you go back to 7 it should be fine.


Sent from my iPhone using Tapatalk

 

Doesn't this issue point at a major problem in the way these displays are manufactured? Surely there is no need for the user ever to be able to reprogram the EDID data in a panel so this feature shouldn't even be possible. I would be interested to hear other peoples thoughts on this but it would be my opinion that the display EDID data should be stored on a read only chip that can only be reprogrammed at LG's or Samsungs or Au Optronics or whichever other screen manufacturers factory with no option to reprogram through the LVDS connector.

 

It's all been answered mate, it's just the thread is that huge unless you've followed you'd have missed quite a lot. The manufactures could have created a WP to the EEPROM but didn't.. doesn't mean they done anything wrong but this is one of the microscopic reasons why we could have done with it. Also it was only until Windows 10 come along and introduced that stupid driver something miniport that software on Windows could alter the EDID, originally this could only really be done through t456 way that being Linux with l2c access. If we are to blame anyone it's Windows 10 for sure, the software that triggers it is likely by accident and the manufacture of this screen could have WP the EEPROM but have you ever known this problem to happen? it's such a microscopic problem (even now only troubling two panels) that I can see why you wouldn't think too much into WP the EEPROM.

 

Why are you not covering the basics first? measure voltage output of the monitor using a multimeter, EDID EEPROM corruption caused by overvolting/undervolting...write protection is useless here. Even overcurrent protection and/or other components ($60 panel) are 'probably' messed up too, turning any previous working driver into a candidate...

Video (incl multimeter), first without NV drivers installed and then install v353.62, bringing back NV to the discussion and forcing them to fix this "officially"...they're using "Third party software" as an excuse.

Spoiler

Bunker edition...

Ubuntu Mate 15.10 Linux kernel v4.2 - Overclocking & Overvolting NV 355.11

1080p full screen.

 

Here's what mine shows... (had to use -i0 for my primary GPU due to SLI or the commands don't work without enumeration)

Code:

C:\[Tools]\NVF>nvflash -i0 --pciblocks

NVIDIA Firmware Update Utility (Version 5.134)

Adapter: N14E-GTX             (10DE,119F,1028,0550) H:--:NRM B:01,PCI,D:00,F:00

The display may go *BLANK* on and off for up to 10 seconds during access to the
EEPROM depending on your display adapter and output device.

Identifying EEPROM...
EEPROM ID (7F,9D22) : PMC Pm25LD020 2.7-3.6V 2048Kx1S, page
Reading adapter firmware image...
Location        Size Type
000000:0171FF  94720 x86
017200:0283FF  70144 EFI

C:\[Tools]\NVF>

My vBIOS dump is attached. I used the code you did... except for the -i0 switch

Code:

nvflash -i0 -d 262144 > eeprom.txt

I sure wish I know how to fix the stupid throttling problem GeForce drivers cause me with anything released after 345.20... maybe some way to block communications between the GPU and the PWRCAPS limits in the BIOS. The goons at NVIDIA say they cannot reproduce the problem, but I can do it on four different systems (M18xR1, M18xR2 - mine and a friend's machine - and the Alienware 18) with 780M SLI. Seems mostly to be limited to 780M SLI and absolutely affected by driver version like flipping a light switch on and off again. But their 'engineers' don't have a clue why or what they changed in drivers after 345.20 that might be causing it. Amazing!

Side note: still no EDID corruption with EVGA Precision X not installed. I am triple-booting W7, W8.1 and Mint 17.2 Rafaela and everything seems to be back to normal again.

I'm about to install 358.20, so wish me luck.

 

As I said earlier I don't have any throttling on 353. However i have 780M stock clocks. Do you experience throttling only when overclocked?

 

I did an update to 358.50. While Nvidia Experience was doing an instalation of the update, my AW18 turned off (crashed, or lost power unexpectedly) - I was scared a bit...
After that I installed them manually without any problems.
LCD EDID survived. Good luck!

 

Pretty sure the throttling issue is when overclocking and not at stock clocks.

 

Nope, that is incorrect... I cannot run any benchmarks at stock clock speeds without throttling to 405 core/400 memory with 780M SLI unless I use drivers 345.20 or older. Just 3DMark11 running stock, test 1 gets about 35-40 FPS. Running stock clocks with 345.20 or 344.75 drivers test 1 is between 70-80 FPS and zero throttling. Same thing happens with Fire Strike and Unigine Heaven. While I can play most games without throttling at stock clocks, doing anything that is heavily taxing with the GPUs, the core and memory clocks drop to 405/400 and it runs like crap on all three of my own machines and the last M18xR2 that I worked on for a customer (also had 780M SLI). Disabling SLI stops the throttling, so there is an additional piece of evidence that incriminates their cancer drivers.

I'm almost the point of being despondent about it. NVIDIA excuses it as being some kind of anomaly, claim they cannot reproduce it, and apparently have no plan to correct it. It's a shame that NVIDIA's incompetence is all it takes to make me want to get rid of the best laptop ever created in the history of computers. Unless I use drivers that are 10 months old, my 780M SLI is utterly worthless to me now... all because of their screwed up drivers and lack of ownership in resolving it. Maybe this is their business model to sell new product, because 780M SLI is still extremely powerful and there are many people content with it, they have to build cancer drivers that retroactively cripple yesterday's flagship GPU and fabricate the sense of need to upgrade or buy a new machine.

 

Anyone up for a round of Micro$loth Roulette? We are one week and a day into my EDID corruption-free life on the M18xR2. So far, so good.

Now that I'm pretty handy at fixing the screwed up LCD panels and eventually figured out Precision X was not playing nice with whatever Windows 10 cancer messed up, I'm seriously considering rolling the dice on @t456 idea of seeing if reinstalling Windows 10 will re-flip the write-protection bit and fix what it originally screwed up. Also wondering if all of this chit-chat going on between Alienware, NVIDIA and the Redmond Mafia if they maybe silently undid whatever they (Micro$lop) messed up and might pretending like it's not their fault.

What I don't like is the possibility that NVIDIA might decide they don't need to help figure this out and try to blame it on EVGA Precision X when some of us already know it did not cause it. If the bricking starts again with Windows 10 and no EVGA Precision X they won't be able to use that excuse if they were planning on trying to.

I think I might roll up my sleeves and dive in to see if anything blows up. What have a I got to lose at this point? 780M SLI that is worthless unless I use 10 month old drivers to avoid throttling? A BIOS that no longer supports GPU upgrades unless I downgrade to pure UEFI? Not much to live for any more as an Alienware owner?

When everything is all said and done, I still have this amazing thing called a "Eurocom Panther" that's to die for awesome. OK, let's go... wish me luck.

As they say in Mortal Kombat... Round 2... FIGHT!

@Matrix Leader

 

I was actually thinking of suggesting you try Windows 10 without Precision X. However I didn't because you definitely seemed like you hated 10 lol. I think it will be a nice experiment for sure.


Sent from my iPhone using Tapatalk

 

Oh, I definitely do hate Windows 10. It sucks. But, it's not going away and the day is coming when they will stop supporting Windows 7. Remember, we're dealing with the Redmond Mafia... they don't care what we want. If I can get it to stop screwing up my LCD panel, then I can focus on butchering the OS and making it what I want it to be like I did with Windows 8 and 8.1, defeat their telemetry/force update cancer and maybe give something of value to the Kool-Aid drinkers that think Windows 10 is a good product. It's not, and may never actually be good. Windows 8/8.1 was, and continues to be, a sucky OS. But, like Windows 8/8.1, we can fix some of what they have left broken and make it better through hacks and tweaks.

I think there was something wrong with the first Windows 10 RTM ISO that killed a lot of our display panels. I cannot prove that, but if there was I would be willing to bet they figured that out on their own and quietly fixed whatever the problem was to avoid more bad press. I am not going to use that ISO again. In fact, I've already deleted it.

 

Getting things ready... got to have my exit strategy available if I need it, LOL. (Obviously, not a committed customer talking here... easy come, easy go.)

 

That's a shame! I am using the 10240 ISO to modify the .wim files. I'll pm a nice till for you if you prefer not using AIK or NTlite. It can remove Cortana and edge (although it supposedly breaks the search bar, but they have IE on it hiding, so removing edge will NOT frak the os, but needs done pre-install for best use). Also plan on removing all hyper-v, defender, and firewall (I use a third party for mine) in mine, the Microsoft store and xbox (PowerShell for that, but get Netflix app before removing those and almost all other apps) are gone. Then, dealing with the registry to add a line to allow the same override as is in enterprise edition. Blocked some connections for telemetry on the router. Disable/delete some services and tasks in task scheduler. After all that, it will begin resembling an os again!!!

I'll give you some good links on the topic when I pm you...

 

Awesome, thank you for that. I appreciate it.

Not crazy about installing this crappy OS, and I'm kind of hoping it bricks my display just so the goons at NVIDIA and Micro$loth cannot use EVGA as a scapegoat. Of course, if the original Windoze OS X ISO was flawed and the current is not, we may actually never get to the bottom of this mess.

Edit: And, of course, installing Windows 10 on my M18xR2 can only be regarded as an experiment because the drivers for Windows 10 are too new to not allow me to use 780M SLI without SEVERE throttling. Windows 10 won't be usable for benching because I cannot use drivers old enough to allow my GPUs to function as they are designed to because NVIDIA has never fixed that problem with their screwed up drivers.

 

It really is too bad that all of the companies that we held to such a high regard are falling into shambles of their once great selves. We need some new companies to form up and clear this filth. At the very least, provide competition so they actually have to try.


Sent from my iPhone using Tapatalk

 

That's odd. What about vBIOS in 780M? Do you also have UEFI->YES ticked in gpu-z?

What if you force 3D clocks in nvinspector?

 

@Mr. Fox @j95

Info from LaptopVideo2Go admin mobilenvidia:

 

Yes, UEFI is ticked in GPU-Z and the vBIOS has UEFI code in it. Maybe an old one with that removed would be better, but I cannot find one. Even the first Clevo and Dell 780M vBIOS is UEFI.

Forcing 3D clocks with Inspector does not help. I've tried that. I have also tried multiple vBIOS and modding my own, but nothing works. It is related to power draw. As soon as it approach the 325W to 330W level, the clock drop immediately. That is why many games do not throttle, because they don't load the GPUs up to draw that much, and neither does a single 780M. NVIDIA must have done something in the drivers to limit power on 780M SLI. Probably some glitch they don't want to take time to look for, because it does not happen with 680M SLI or 880M SLI.

I am not sure what to make of that. I am no longer having the EDID corruption now with Windows 10 (so far, but we will see in a day or two) and no EVGA Precision X. The 358.50 drivers throttle really bad in Windows 10 also, same as Windows 7. I am running my modded 345.20 with Windows 10 and it is working fine. I tweaked 344.75 for Windows 10 and it installs and works fine except there is no option to enable SLI. Windows 10 is installed in pure Legacy BIOS (no UEFI with Legacy Option ROM).

 

I had also the Problem with the EDID corruption with W10 (Since W10-Launch installed). I installed EVGA Precision X and wanted to restart. Got 8 beeps.... :/ . So I had to restore my EDID without problems. After that i uninstalled drivers (355.98) + Precision X, installed 353.17 (working great without throttle). Also no more corrupt EDID for 7 days yet. So i would say that noone should try to do install Precision X till there is a fix or sth.

 

I can't PM you. Any way you can start a conversation with me to send you the info on modifying Win 10? Unless you want me to post it here...

 

Sorry, I sent a message you should be able to reply to. I need to turn it back on again.

---

Spoiler: Day 2 Update

And, for Micro$loth...

Spoiler: A private message...

 

Haha, good to know that it seems almost definitely that it was Precision X that was the trigger for the screen bricks (albeit with the proviso that Windows 10 was installed at some point) and that you're surviving on Windows 10 for now! I suppose it's yet to be found what the other trigger is for screen bricks for the one or two people that never installed Precision X.

 

Yes, interesting indeed. I hope the original trigger is found. As a number of others posted, their screens got corrupted without EVGA Precision X. In the case of my Alienware 18, it was not installed. After switching from Intel HD Graphics to NVIDIA and installing GeForce drivers it bricked immediately. I was not even totally finished installing my software yet, LOL.

We should not overlook the fact that I am using a different build and there is a possibility Micro$loth quietly replaced the first available RTM ISO that I used. Something may have actually been wrong with the product build that I and others downloaded that is no longer present in the current ISO. If that is the case, the root cause may never be identified and it may never be possible to hold them accountable for their mistake.

 

Ok, new display on its way

Did anybody successfully use the 1GB usb version from t456? Or should I go with the original 32gb one?

i watched some m18xr1 tear down videos. Do you have any tips for me regarding how to set up for the hot swap?

Thanks again.

 

Buy another LVDS cable if one is not coming with your new display. Leave it connected to the new display indefinitely and use the new display only for hot-swapping to fix the old one. This will save you time having to take apart the old screen for no good reason. It's definitely worth the cost of the LVDS cable. Don't bother installing the new LCD. That would just be a waste of time, plus you'll want to keep it pristine (unless your old screen is scratched up or something like that) for future use or resale.

Just take off the bezel around the keyboard. No need to go any deeper than that. Use a USB keyboard and mouse. Take the six screws out of the keyboard and TactX keys and gently lift their tabs out of the palm rest (slide the keyboard toward the screen) and gently lay them face down on the palm rest. No need to disconnect the ribbon cables, just be gentle and don't pull on them.

• Set the new LCD in front of the bricked one on the area by the hinges. It should stay propped up there fine, leaning against the screen behind it.
• Disconnect the old LVDS cable, connect the new LVDS cable and boot up Linux.
• Once you are on the Linux desktop, figure out the commands, dump your EDID.bin from the new display, etc.
• When you're ready to flash and have figured out exactly what to do, unplug the new display (machine still running) and put it back in the box or another safe place.
• Connect the original LVDS cable. Be sure to get rid of any ESD build-up by touching something metal to ground yourself to first, like one of the hinges, to eliminate any ESD.
• The bricked display will be working fine at this point, appearing as though nothing is wrong with it. And, it does work fine, it just won't POST with corrupted EDID.
• Flash it with one of the good EDID.bin files @t456 provided for your display or the one you just dumped from the new display. Watch the terminal window when you flash it and make sure the writing process completes without errors.
• Reboot into Windows and make sure everything is working. Shut down and put it back together.

Once you have done it a couple of times you can fix your M18xR2 beast in probably 10 minutes or less. I can just about do it with my eyes closed now... not quite, but almost.

 

Finally found the article I thought of when this started, with the firmware tie in:

Dell apologises for leaving NSA backdoors in hardware
http://www.kitguru.net/gaming/secur...ogises-for-leaving-nsa-backdoors-in-hardware/

Earlier this week at the 30th annual Chaos Communication Congress, Dell, among other US based tech firms, was accused of leaving deliberate backdoors in its software and hardware, to make it easier for the NSA to install malware and viruses on the machines and gain access to the information stored on them. This affects consumer hardware as much as enterprises, suggesting that many of the world’s computers could be lifting their digital skirts to the US for less than silver dollar. In response to consumer concerns about this, Dell has apologised, poorly.

“Thanks,” said one tech-head via a Tweet to Dell earlier this week. “I just found out my Dell server has NSA bug in Rand (RAID?) BIOS.”

Dell’s response?

“Thank you for reaching out and regret the inconvenience,” it said. “Our colleagues at @DellCaresPro will be able to help you out.”

Speaking of hardware with RAID firmware...

Dell to acquire EMC for $67 billion- largest deal in tech history
http://www.kitguru.net/channel/gene...-for-67-billion-largest-deal-in-tech-history/

"According to Dell’s own post on the matter, this will bring together the industry’s leading innovators in the data centre field. This deal will also have the added benefit of turning Dell in to an “enterprise powerhouse”, making this a pretty major move.

For years, a significant chunk of Dell’s business has relied on the sales of PCs, though it has been moving in to the enterprise sector with new devices, like its upcoming Surface competing tablet. Now with EMC under its belt, Dell will have a much larger presence in enterprise, cloud computing and data centre business, Dell will also be acquiring EMC’s 81 per cent stake in VMWare, which accounts for a large portion of the company’s value."

 

Good old Micro$haft pushing updates that should be a choice.
http://news.softpedia.com/news/micr...upgrade-enabling-data-collection-494236.shtml


Sent from my iPhone using Tapatalk

 

They are so stinking evil... and, the'll say, " What's wrong, sweetie... you know you really like this, admit it," as they are virtually punching us, kicking us, knocking out our teeth, and pulling our hair out in fistfuls. People deserve to go to jail for this kind of thing.

 

Is it possible to summarize what EVGA Precision X version the majority used along with Win10/updates in 7/8 before screen bricks? It should also have been possible to create a poll on which Evga precision version people used in this thread.
Has anyone been in touch with EVGA forum and talked about this possible issue? Not saying that EVGA precision is the problem but you never know with this software + Windows 10.

 

is.this problem of lcd bricking only with alienware

what about.msi and asus

Sent from Galaxy Note 5 via Tapatalk

 

It's just windows 10 opened a hole after being installed and there are some software which would accidentally use that bug for something which was not intended. EVGA precision might be the one. I bet that Nvidia drivers had something inside too and now perhaps it's removed, perhaps not.
It's my educated guess, of course, no facts.

 

So let's do a summary:

• Nvidia got reports
• Nvidia after 1,5 months and tens of destroyed displays finally found bug
• But they said that it's dell/clevo bug (wtf?)
• They sent it to dell(?)
• After weeks, still no answer from dell and nvidia both. Laptops display are still being destroyed.

Maybe we should start a spam crusade on dell forum?

 

still unsure if not having precision x installed on a win 10 install will break the screen. not risking it for my 120hz panel.

 

I think your assumption here is very close to reality. This is my theory as well. Something that Windows 10 did is the root cause, NVIDIA drivers and EVGA Precision X do things that are not good once the hole gets opened. It didn't matter before because there was no hole. I'm not sure Windows 10 is still punching the hole though. They may have quietly fixed that without ever admitting there was a problem.

 

First of all i would like to thank t456 and Mr. Fox, second of all i would like to thank all of you that contributed to this topic.
I manage to restore the edid's to my 2 messed up lcds.
The work was not hard with extra care, but not easy as well.
I'll come back with more details later on.
Thank you all.

 

If they fixed it, they didn't change the build number. Media creator downloads a 10240 build. I need to compare a couple things between the esd and wim to check for changes (other than the baked in updates, which are listed in ntlite free version)... I'll throw in one of my Sata drives hoping it is there and not on my m.2 (otherwise it will take longer for confirmation). I converted it to an ISO also previously, but would have to find that too...

 

If anyone here is part of the insider program they can install the latest build which I believe is 10532. They may be releasing the newest build this week also which is 10565. Maybe you can see if that hole is still there. I don't have my new AW 18 yet so there is no way for me to check.

 

Just check md5

 

I was wrong. Latest release is 10547. I can't check the md5 until I get home tonight.

 

Add me to the list of those with a bricked LCD. Below are some details.

Alienware 18 running Windows 7 with W10 ready updates (the system has never seen anything but W7). sBIOS is A12
NV 980M with svl7 mod bios (previously 780M with svl7 mod bios)
Old version (don't know version number) of Precision X followed by update to latest (as of 12OCT2015 Precision X)

History:
I was running W7 with W10 ready updates installed with the 780M and modded vBIOs with the latest drivers (355.82 at the time) and an old version of Precision X for about 2 months without any problems.

About a week ago I upgraded to the 980M did a clean driver install via DDU and installed 353.06 from T|I. Kept old version of Precision X. This configuration ran fine for about a week.

Fast forward to the weekend and I installed 358.50 WHQL. Updated to latest version of Precision X. Everything worked fine for that evening. I wake up yesterday morning to 8 beeps. Tried everything to get it back up, swap video cards, run without MXM cards, ect. Nothing helped.

I just ordered a new LCD screen and we will see how it goes.

Once I get the new panel installed I plan on doing a fresh install of W7 and disable any updates. I will also not be installing Precision X (neither the old version nor the updated one). Since I had good luck with 353.06, I'll stick with that. Wish me luck and we shall see how it goes.