Optimus Pro Xi - AMD / Nvidia 2060 - Unusable due to Blue Screens but it’s not an hardware problem

[DragonDog]

Hello everyone, in August I bought a 17.3 "OPTIMUS PRO XI with a ryzen, nvidia 2060 and a 1tb intel SSD. The computer arrived and after a few weeks the problems started. Every now and then the disk was not recognized at the bios level (black screen with white text that says that there was no disk inside the pc). After several attempts (on and off) the pc turns on.

The most serious problem is the one related to blue screens. More or less every 2 days a blue screen appears and it’s impossible for me to use the pc for work (which is why I bought this pc).

These are the last bluescreen control strings I encountered in the last few weeks all caused by the driver "ntoskrnl.exe" and the address "ntoskrnl.exe 3f5210"

IRQL_NOT_LESS_OR_EQUAL
KMODE_EXCEPTION_NOT_HANDLED
IRQL_GT_ZERO_AT_SYSTEM_SERVICE
KMODE_EXCEPTION_NOT_HANDLED
DRIVER_IRQL_NOT_LESS_OR_EQUAL

I have formatted multiple times and even after formatting it keeps giving me the same problem. Many times after the pc stops due to a bluescreen on the next reboot the 1tb Intel SSD is not even recognized and I am forced to restart the pc again.

This is the list of all the attempts I have made:
1) Installation of all the drivers provided in the download section (pc specialist website)
2) Formatting and installing drivers directly from Windows Update
3) Formatting and reinstalling the drivers provided by pc specialist website
4) Uninstalling integrated video card driver (later dedicated)
5) Control via sfc / scannow
6) Several memtest
7) I have tried softwares that check the health of the disks and the memories and nobody detects anything
8) I called the phone support who tell me to try to use the pc with a different ssd, but after a couple of days bluescreen again
9) The assistance asked me to also check the disks through the official Intel tool "Intel SSD Toolbox" but even in that case no errors.
10) I also tried to update the SSD firmware

I no longer know what to do, plus the bluescreen do not occur every day, so I have to continually use the PC with the fear that it will turn off making me lose all the data I am working on.

The Windows ISO has been downloaded through the official channel each time and the same iso has never been used to avoid the possible problem of corrupted Windows ISO.

From today, however, the pc is completely unusable and I cannot use it for more than 20 minutes straight before it gives me a blue screen.

I thought it was now a hardware problem but still today (before opening a ticket to send it to assistance) I decided to use linux in live to recover my work files from the SSD since windows was not usable for more than 10 minutes without a blue screen , and... with linux the pc works without problems. At this point I really think it's a driver problem, but I'm not 100% sure.

Have any of you had a similar problem? Can you tell me something about it? Thanks

 

[ubuysa]

Hello and welcome. I'm sorry you're having these problems.

The key sentence in your post is this one...

with Linux the PC works without problems

...because that suggests that the hardware is fine (although it is still just possible that Linux isn't using some advanced hardware feature that is the problem). Based on this we can concentrate on getting you a stable Windows platform.

When you say that 'you formatted the PC' how did you do that? Was this a fully clean install from bootable media? If it wasn't then you need to do a fully clean install...

1. First backup all your user data.
2. On another PC (because you can't trust this one) download the Windows Media Creation Tool. Plug in a blank 8GB (min) USB stick and run the tool. Tell the tool to create installation files for use on another computer and point it at your USB stick. You'll then have a bootable USB stick containing the Windows installation files for the latest version of Windows 10 (version 20H2).
3. On your PC boot the USB stick containing the Windows installation files (you may need to change to boot order in the BIOS setup to do this).
4. Click the big blue Install button and then select a Custom Install.
5. Delete all four UEFI partitions on your system drive (your SSD) - these will be labelled MSR Reserved, EFI System, Recovery, and Primary - so that all that space shows as unallocated space.
6. Click (to select) that unallocated space so that it's highlighted and then click the Next button. The installer will create the correct partition structure and install Windows.
7. When Windows is installed you'll need to run through the initial setup (language, region, userid, etc.).
8. Once the initial setup is done run Windows Update repeatedly (even across reboots) until no more updates are found. This will install all necessary drivers. You may want to visit the Nvidia website to see whether there is a later driver for your 2060.
9. Open Device Manager and check that there are no devices shown with a yellow triangle (with an exclamation mark in it) next to them. If there are none shown then your drivers are good.

You now have a stable Windows platform, so before installing anything else or making any other configuration changes, test it as thoroughly as you can to try and make it BSOD. If it does BSOD in that state it's probably a hardware problem. If it does BSOD look for minidumps in the C:\Windows\Minidumps folder and a kernel dump in the file C:\Windows\Memory.dmp. Upload any of these you find to the cloud somewhere and post a link here, I may be able to isolate what the problem is.

 

[DragonDog]

Thank you for the support! I formatted the pc exactly with this method, the only thing that the various installations I have done have in common is the fact that in all of them I have updated windows through windows update. With Linux I had no problem but it was not installed, it was live from usb (so maybe the problem is the main SSD?) even if I have done tests with some tools and nobody finds problems in the disks. I leave the dump files in this dropbox link. You can find both the blue screen dmp and the memory dump. I've only included some of the latest .dmp files. Thank you for the support!

Dropbox

www.dropbox.com

 

[ubuysa]

The minidumps first then...

112420-8187-01.dpm is unusual, it's an IRQL_GT_ZERO_AT_SYSTEM_SERVICE. This is caused when a thread returns to user processing mode whilst it's IRQL is above that used for user mode code. The process in control at the time of the error was CEPHtmlEngine.exe, which is part of Adobe Photoshop.

112620-7796-01.dmp is a KMODE_EXCEPTION_NOT_HANDLED. This is caused by a kernel mode process encountering an error that the error handlers couldn't handle. In this case it was a STATUS_ACCESS_VIOLATION (0xC0000005) which is a memory access violation, the address that was attempting to access is invalid (0xffffffffffffffff) which looks like a driver problem. The process in control was System.

113020-7421-01.dmp is a common IRQL_NOT_LESS_OR_EQUAL which happens when a driver accesses a memory page that is either not valid or allocated. The process in control was dwm.exe, which is the Desktop Window Manager. There's no indication of the driver involved but in the stack trace just prior to the page fault are some calls to I/O handlers so it was related to reading or writing to a device (possibly the SSD).

120720-111-71-01.dmp is also an IRQL_NOT_LESS_OR_EQUAL and for exactly the same memory location as the one above (0x2000). The process in control is different though, this one is System. The stack trace here shows a nt!memset call (a memory access) just prior to the page fault.

The kernel dump is also an IRQL_NOT_LESS_OR_EQUAL and for exactly the same memory location as the one above (0x2000). It is in fact identical to the previous minidump (including structure addresses). It's not uncommon to find the kernel dump and the latest minidump are identical, they were both written for the same event by different components.

So....what have we learned?

These do all look driver related, but there doesn't seem to be a common owning process nor indications of a common driver, so that might be a red herring. Perhaps.

RAM could well be an issue, a couple of the dumps do point at RAM but all of them could have a RAM cause. I see you've run 'several Memtests', were these the Windows Memory Diagnostic or the (better) one from https://www.memtest86.com? If you've not run the Memtest in that link then please do so, it's far more thorough than the Windows one. It will take several hours to run so leave it running overnight - let it complete all four iterations of the various tests.

Even if Memtest runs clean, and if you have more than one RAM stick in your laptop, it's well worth running with just one stick at a time for long enough to tell whether it's going to BSOD or not. Try each stick in turn to see whether you can get it to BSOD with one of them. This will also allow you to reseat each RAM stick which is another useful thing to do.

If the SSD is an M.2 SSD then undo the locking screw, remove the SSD and replace it, and the locking screw. We've had a couple of people recently solve curious issues by reseating the SSD.

Having Linux run of the USB stick does give most components a good test but it would be much more reliable to install it onto your SSD to confirm that it will work OK running internally, see whether it will fail at all then.

My gut feel based on everything you've done and the dumps here are that this is most likely to be a hardware issue, some feature of some device that Windows uses and Linux doesn't, or which Windows uses in a different way to Linux. The fact that you have clean installed Windows (several times!) and allowed Windows Update to install all drivers and it still BSODs is a pretty clear pointer to a hardware issue.

BUT. When you clean install Windows and run Windows Update it is essential to install no other software nor make any configurations changes before you test it. You need the most stable and reliable Windows platform you can get for testing purposes, this only comes from a clean install of Windows AND NOTHING ELSE.

 

[DragonDog]

First of all, thank you very much for your interest and for the very detailed reply. I used the windows integrated memtest, but tonight I will leave the pc to do the test you recommended (memtest86).

Meanwhile, I have some news! But bad news... I reinstalled Windows from scratch (fresh install, new iso again and all driver installed by windows update) but this time not in the 1tb intel ssd but in the secondary disk I installed. I have anything installed but the basic essential software. After about 4 hours of use ... everything froze. But no blue screens this time and no minidumps. I'm really losing hope.

Now I try (as recommended by you) to remove the SSDs and put them back. Your idea of keeping the pc empty only with windows installed and its drivers is certainly valid, the fact is that I bought this pc to work, and at the moment I don't have a second pc. I'm in smartworking and I really can't afford to wait for a freeze or a blue screen! I think you can understand! 😢

Tomorrow when the test is finished I will write the results of the memtest86

Also the notebook is just 4 months old, so it's practically new. That's very sad.

Thank you again! 😊

 

[DragonDog]

Tonight I started MemTest86 and once again zero errors ... I left the log file and the screenshot of the final test result inside the Dropbox link. I think it's more of a software problem than a hardware problem, probably driver compatibility. Link to dropbox folder https://www.dropbox.com/sh/kk4oqptla6sbule/AACGqbGJtX49K0jrj3YnhpUFa?dl=0

Anyway now I try to remove and put back the SSDs. Today I work all day with the pc, so I can check if there are any problems.

 

[ubuysa]

Tonight I started MemTest86 and once again zero errors ... I left the log file and the screenshot of the final test result inside the Dropbox link. I think it's more of a software problem than a hardware problem, probably driver compatibility. Link to dropbox folder https://www.dropbox.com/sh/kk4oqptla6sbule/AACGqbGJtX49K0jrj3YnhpUFa?dl=0

Anyway now I try to remove and put back the SSDs. Today I work all day with the pc, so I can check if there are any problems.

Can you post the full spec of your build from the original order please so we can all see exactly what your build looks like?

 

[DragonDog]

Optimus Series: widescreen LED 72% NTSC, full HD 17,3" - 144 Hz (1920 x 1080)
AMD Ryzen™ 7 4800H 8 core (2,9 GHz, 4,2 GHz Turbo)
DDR4 SODIMM Corsair 2666 MHz 16 GB (2 x 8 GB)
NVIDIA® GeForce® RTX 2060 – RAM video 6,0 GB GDDR6 – DirectX® 12.1
SSD PCIe INTEL® 660p M.2 NVMe 1 TB (1800 MB/sR | 1800 MB/sW)
1 Power Adapter CA 180 W
Bluetooth - wireless LAN GIGABIT | WIRELESS INTEL® Wi-Fi 6 AX200 (2,4 Gbps) + BT 5.0
USB/Thunderbolt 1 USB 3.1 (type C) + 2 USB 3.0 + 1 USB 2.0

I only added another 256GB SSD in the in the secondary pci slot

 

[ubuysa]

Ok. There are two things I'd suggest you try (one at a time though)....

1. Remove the extra SSD you added for a few days. I do know that some laptop models have some issues with an SSD in that second M.2 slot (there was one the other day that would not boot from it for example). When things start to fail it's always wise to back out the thing you changed last - PCS will want you to do that in any case if you ask them for help.

2. Remove one of the RAM sticks for a few days. I know Memtest produced no errors, but if you look closely at the log a lot of memory locations were ignored - especially in test 13; the hammer test. In any case, Memtest can only show that RAM is definitely bad, it can't ever show that RAM is definitely good. Running with one RAM stick is the best way to eliminate RAM. Swap the sticks over after a few days.

Is there anything using the Thunderbolt port? If there is (and you can live for a few days without it) I'd unplug that too.

A do agree that all the dumps I've seen are typically driver problems but I don't see much linkage between them to point to a single driver, and if you're doing fully clean installs (deleting existing partitions) and using only Windows Update drivers (and the latest Nvidia driver from their website) then it's hard to imagine that this isn't a hardware issue.

Can you also download HWMonitor and post your temperatures - both at idle and as loaded as you can make it? I want to be sure this isn't an overheating issue....

I would suggest you start to try and work out a procedure that always makes it BSOD - or which has the highest probability of a BSOD - because when it gets to the PCS test bench (if you end up having to RMA it) they have to know how to make it fail. If they can't see it fail they're unlikely to be able to fix it.

I'm going to spend some time diving a bit deeper into those dumps.....

 

[bengro]

I'm interested in this post as well. I have the same spec laptop with the differences being 32gb Ram and Sabrent 1tb Nvme. I get a BSOD every couple of days - but always on boot up. I suspect it is the nvme and have purchased a replacement, but as yet to install it - although an updated bios might be the best solution. I have used memtest to check the ram and that passed with flying colours. Temperatures are all within expected limits etc and the laptop is always on a cooling base.

 

[ubuysa]

I'm interested in this post as well. I have the same spec laptop with the differences being 32gb Ram and Sabrent 1tb Nvme. I get a BSOD every couple of days - but always on boot up. I suspect it is the nvme and have purchased a replacement, but as yet to install it - although an updated bios might be the best solution. I have used memtest to check the ram and that passed with flying colours. Temperatures are all within expected limits etc and the laptop is always on a cooling base.

You're very welcome to follow along here of course, but if you need help yourself please start a new thread - otherwise it gets too complex!

 

[DragonDog]

Ok. There are two things I'd suggest you try (one at a time though)....

1. Remove the extra SSD you added for a few days. I do know that some laptop models have some issues with an SSD in that second M.2 slot (there was one the other day that would not boot from it for example). When things start to fail it's always wise to back out the thing you changed last - PCS will want you to do that in any case if you ask them for help.

2. Remove one of the RAM sticks for a few days. I know Memtest produced no errors, but if you look closely at the log a lot of memory locations were ignored - especially in test 13; the hammer test. In any case, Memtest can only show that RAM is definitely bad, it can't ever show that RAM is definitely good. Running with one RAM stick is the best way to eliminate RAM. Swap the sticks over after a few days.

Is there anything using the Thunderbolt port? If there is (and you can live for a few days without it) I'd unplug that too.

A do agree that all the dumps I've seen are typically driver problems but I don't see much linkage between them to point to a single driver, and if you're doing fully clean installs (deleting existing partitions) and using only Windows Update drivers (and the latest Nvidia driver from their website) then it's hard to imagine that this isn't a hardware issue.

Can you also download HWMonitor and post your temperatures - both at idle and as loaded as you can make it? I want to be sure this isn't an overheating issue....

I would suggest you start to try and work out a procedure that always makes it BSOD - or which has the highest probability of a BSOD - because when it gets to the PCS test bench (if you end up having to RMA it) they have to know how to make it fail. If they can't see it fail they're unlikely to be able to fix it.

I'm going to spend some time diving a bit deeper into those dumps.....

Ok, as you recommended first I removed the SSD installed by me. I also swapped the rams and took out and put back the main drive (1tb Intel SSD) as well. I have to work with the notebook this afternoon, so I can check that everything is fine, I'll update you tonight! Thank you!

 

[ubuysa]

Ok, as you recommended first I removed the SSD installed by me. I also swapped the rams and took out and put back the main drive (1tb Intel SSD) as well. I have to work with the notebook this afternoon, so I can check that everything is fine, I'll update you tonight! Thank you!

OK. I spent a couple of hours with those dumps and I'm not able to either isolate a specific driver at fault nor find a driver connection between them. I even went through the list of installed modules in the failing thread in each dump and none of them stood out as unusual.

I don't think this is a software or driver issue....

 

[DragonDog]

Here I am, thank you so much for the help you are giving me. I was working all evening without problems when just a little while ago a blue screen appeared. But I didn't have time to read the error message, it remained for about 1 second and then the pc shut down and restarted by itself. No minidump

At the moment there is only intel SSD of the initial configuration and I swapped the ram banks. I fresh installed windows yesterday.

I don't know what to do anymore, I'm wasting too much time with this problem, I think the only solution is to make a rma. I am very disappointed with this pc.

Tomorrow I try to work with only one ram bank as you advised me, but I will contact the assistance for the rma.

 

[ubuysa]

Here I am, thank you so much for the help you are giving me. I was working all evening without problems when just a little while ago a blue screen appeared. But I didn't have time to read the error message, it remained for about 1 second and then the pc shut down and restarted by itself. No minidump

At the moment there is only intel SSD of the initial configuration and I swapped the ram banks. I fresh installed windows yesterday.

I don't know what to do anymore, I'm wasting too much time with this problem, I think the only solution is to make a rma. I am very disappointed with this pc.

Tomorrow I try to work with only one ram bank as you advised me, but I will contact the assistance for the rma.

Definitely try one stick of RAM at a time - I still think that may be the problem. I would leave the second SSD out for now as well.

Have you noticed any particular pattern when it BSODs? It is only when using a specific app or game? Is it only when using a particular mix of apps? Is it only when using a specific device? If you do eventually RMA it you need to give PCS instructions on how to make it BSOD. If it won't BSOD on their bench they will just ship it back to you.

You want to stop Windows auto-restarting on a BSOD so that you can read the error message. In the Run command box enter sysdm.cpl and click the Advanced tab. In the Startup & Recovery section at the bottom, click the Settings tab and uncheck the box that says Automatically restart and click OK. If a BSOD happens now the system will halt on the blue screen and you'll have to manually restart.

 

[ubuysa]

I've just spent some more time on your kernel dump - the minidumps (as their name suggests) only contain details of the failing process and thread, whilst the kernel dump contains the whole kernel and related control areas, so there's much more detail in there.

On closely checking the loaded modules there are several there for VMWare (vmci.sys, vsock.sys, vmkbd.sys, vmnet.sys, vmnetadapter.sys, hcmon.sys, vmx86.sys, vmnetbridge.sys, vmnetuserif.sys). Are you running VMWare at all? If so what guests you you have running in there? Do you get these BSODs with VMWare stopped? Better yet, do you get these BSODs with VMWare uninstalled?

 

[DragonDog]

Hi, since last night I have been using only one ram bank, and for now still no blue screen. Unfortunately there is no particular pattern, they are completely random, both when I use some software and when I have everything closed or when I play games or surf the internet. Sometimes even when I leave the pc with everything closed and don't use it. But as I said they don't have a pattern, so this makes it difficult for me to explain the problem to PC Specialist support.

However, although I have not yet encountered blue screens since I am using only one bank of ram I still cannot be sure that the problem is fixed because (as I said) the BSODs do not have a pattern!

For now, I have followed the procedure you described to not restart the PC after a blue screen. Thanks again for analyzing my files!

After formatting 2 days ago I did not reinstall VMWare, and despite this the notebook froze and did a blue screen anyway (the ones I already reported to you in the past few days!). I continue to use the pc with only one ram bank, and in a few days, if all goes well, I change the bank again and see if the other one gives me the problem.

 

[ubuysa]

Hi, since last night I have been using only one ram bank, and for now still no blue screen. Unfortunately there is no particular pattern, they are completely random, both when I use some software and when I have everything closed or when I play games or surf the internet. Sometimes even when I leave the pc with everything closed and don't use it. But as I said they don't have a pattern, so this makes it difficult for me to explain the problem to PC Specialist support.

However, although I have not yet encountered blue screens since I am using only one bank of ram I still cannot be sure that the problem is fixed because (as I said) the BSODs do not have a pattern!

For now, I have followed the procedure you described to not restart the PC after a blue screen. Thanks again for analyzing my files!

After formatting 2 days ago I did not reinstall VMWare, and despite this the notebook froze and did a blue screen anyway (the ones I already reported to you in the past few days!). I continue to use the pc with only one ram bank, and in a few days, if all goes well, I change the bank again and see if the other one gives me the problem.

Ah that makes sense. I was under the impression the dumps were from a vanilla Windows install. [emoji3526]

Do give it a few days on that RAM stick if you can, you need to be certain it's not failing before swapping to the other stick.

Please upload any dumps that are generated if it does BSOD.

 

[DragonDog]

Meanwhile, I have some news! But bad news... I reinstalled Windows from scratch (fresh install, new iso again and all driver installed by windows update) but this time not in the 1tb intel ssd but in the secondary disk I installed. I have anything installed but the basic essential software. After about 4 hours of use ... everything froze. But no blue screens this time and no minidumps. I'm really losing hope.

Yes, those dump files were the ones present in the previous Windows installation. Then (as I wrote you in this answer) I have formatted and done a new clean install!

For the moment I am still with only one bank of ram and still no blue screen. 😀 I think the ram could be the cause of the problem. I keep it for a few more days and then I change with the other the bank!

Of course, if it does other blue screens I will load the updated dump files!

 

[ubuysa]

Yes, those dump files were the ones present in the previous Windows installation. Then (as I wrote you in this answer) I have formatted and done a new clean install!

For the moment I am still with only one bank of ram and still no blue screen. [emoji3] I think the ram could be the cause of the problem. I keep it for a few more days and then I change with the other the bank!

Of course, if it does other blue screens I will load the updated dump files!

It would not surprise me to find that each RAM stick is fine on their own but BSOD when you use both together. This is such a strange issue that it would not surprise me to find it's some sort of dual-channel issue and that the two sticks, although on the face of it a matched pair, have some sort of minor difference that is tripping you up? We'll see....