Baffling Palit 2080Ti problem

[laserom]

Hi, my PC randomly crashes while general internet browsing or while gaming.
It would stall then restart. No blue screen, just a straight crash.
I have a temp monitor on the 2nd screen so I can see immediately what the temps were at time of crash. At every crash the 2080Ti has not gone above 40 Celsius.

I have DDU'd at least twice, applied different drivers in attempt to resolve it (Dec, Jan, Feb and March Nvidia drivers.)

If anyone is wondering my PC is this https://www.overclockers.co.uk/ocuk...700k-4.6ghz-geforce-rtx-2080ti-fs-1b8-og.html

Below is the report from WhoCrashed app. (It actually crashed yesterday, 8/3/2020, but for some reason a dump wasn't generated.)

Any ideas?

>>>

Crash dumps are enabled on your computer.

Crash dump directories:
C:\Windows
C:\Windows\Minidump

On Mon 09/03/2020 16:39:24 your computer crashed or a problem was reported
crash dump file: C:\Windows\Minidump\030920-7921-01.dmp
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF801413D897A)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF801413D897A)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL



On Mon 09/03/2020 16:39:24 your computer crashed or a problem was reported
crash dump file: C:\Windows\MEMORY.DMP
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF801413D897A)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF801413D897A)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL



On Sun 23/02/2020 22:14:56 your computer crashed or a problem was reported
crash dump file: C:\Windows\Minidump\022320-9750-01.dmp
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF807225D8ACA)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF807225D8ACA)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL



On Wed 19/02/2020 20:09:01 your computer crashed or a problem was reported
crash dump file: C:\Windows\Minidump\021920-7687-01.dmp
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF80389298BDA)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF80389298BDA)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL



On Tue 18/02/2020 23:45:32 your computer crashed or a problem was reported
crash dump file: C:\Windows\Minidump\021820-9890-01.dmp
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF8067DC58BDA)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF8067DC58BDA)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL



On Mon 10/02/2020 19:31:56 your computer crashed or a problem was reported
crash dump file: C:\Windows\Minidump\021020-10265-01.dmp
This was probably caused by the following module: nvlddmkm.sys (0xFFFFF8036D898BDA)
Bugcheck code: 0xD1 (0x0, 0x2, 0x1, 0xFFFFF8036D898BDA)
Error: DRIVER_IRQL_NOT_LESS_OR_EQUAL
file path: C:\Windows\System32\DriverStore\FileRepository\nv_dispi.inf_amd64_33895c186dfc2a0d\nvlddmkm.sys
product: NVIDIA Windows Kernel Mode Driver, Version 442.50
company: NVIDIA Corporation
description: NVIDIA Windows Kernel Mode Driver, Version 442.50
Bug check description: This indicates that a kernel-mode driver attempted to access pageable memory at a process IRQL that was too high.
This bug check belongs to the crash dump test that you have performed with WhoCrashed or other software. It means that a crash dump file was properly written out.
A third party driver was identified as the probable root cause of this system error. It is suggested you look for an update for the following driver: nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation).
Google query: nvlddmkm.sys NVIDIA Corporation DRIVER_IRQL_NOT_LESS_OR_EQUAL

<<<


Conclusion
6 crash dumps have been found and analyzed. A third party driver has been identified to be causing system crashes on your computer. It is strongly suggested that you check for updates for these drivers on their company websites. Click on the links below to search with Google for updates for these drivers:

nvlddmkm.sys (NVIDIA Windows Kernel Mode Driver, Version 442.50 , NVIDIA Corporation)

 

[Joxeon]

Are you able to test the Gpu in another machine or have access to another Gpu to test in your PC?

 

[laserom]

Are you able to test the Gpu in another machine or have access to another Gpu to test in your PC?

Yeah, I'll try that.

 

[Woodsta888]

That's an overclock crash. Try lowering the overclock with afterburner even if it's at stock boosting levels.

 

[laserom]

That's an overclock crash. Try lowering the overclock with afterburner even if it's at stock boosting levels.

It is a Palit 2080Ti Pro OC and the overclocks are ridiculously high, in the range of 1900MHz, I've not manually overclocked it. I tried to lower the overclock via Precision X1, seems it is locked. Is it MSI AB you're talking about?

 

[opethdisciple]

Yes. My guess is the overclock isn't stable.

 

[Grim5]

Thats a graphics driver crash, often caused by unstable clocks as mentioned. What you should do is try lowering the clocks until the card becomes stable. Once it's stable, check if those clocks are less than what the card has out of the box, if so RMA time.

I would RMA it anyway, once these things become unstable they tend to keep degrading

 

[Woodsta888]

It is a Palit 2080Ti Pro OC and the overclocks are ridiculously high, in the range of 1900MHz, I've not manually overclocked it. I tried to lower the overclock via Precision X1, seems it is locked. Is it MSI AB you're talking about?

Yeah knock off 100Mhz the core and the memory. Try it and if it stays stable. The factory boost for that card is 1575Mhz and it'll probably on it's own boost higher like you are seeing. It just cant hold it anymore. Wonder if OCUK applied an overclock to the GPU, fine during stability testing but wasn't conservative enough and now just can't hold it?

You should be able to access it through precision X1 unless OCUK set it up to lock their OC or stable (at the time) system. Don't bother trying to run afterburner if precision X1 is being used. Precision X1 is EVGA OC'ing software so maybe OCUK use that on all cards as the techs prefer it. Or the one that built your machine does. 40 degrees is very cool for an aircooled card if it's running at max. Would be in the 70+ degrees plus range if running at 100%. What resolution you gaming at?

 

[Grim5]

Yeah knock off 100Mhz the core and the memory. Try it and if it stays stable. The factory boost for that card is 1575Mhz and it'll probably on it's own boost higher like you are seeing. It just cant hold it anymore. Wonder if OCUK applied an overclock to the GPU, fine during stability testing but wasn't conservative enough and now just can't hold it?

You should be able to access it through precision X1 unless OCUK set it up to lock their OC or stable (at the time) system. Don't bother trying to run afterburner if precision X1 is being used. Precision X1 is EVGA OC'ing software so maybe OCUK use that on all cards as the techs prefer it. Or the one that built your machine does. 40 degrees is very cool for an aircooled card if it's running at max. Would be in the 70+ degrees plus range if running at 100%. What resolution you gaming at?

Actually the stock boost clock is 1650 MHz and only has a power limit of 260w.
I suspect OP is confused though, I doubt the card is running with +250mhz offset on the core. Only cards I know that achieve those numbers is the EVGA Kingpin with a 500w BIOS and others at that level.

He's probably just refering to the clock he sees in game, in which case 1900mhz is perfectly normal for a cheap 2080ti if your temp is only 40c. Chances are to get it stable he needs to drop the the boost to under 1650mhz. Just RMA the card.

It can be that 1650mhz is the minimum allowed in the BIOS which is why he maybe can't reduce it.

Again, just RMA the card.

 

[Woodsta888]

Actually the stock boost clock is 1650 MHz and only has a power limit of 260w.
I suspect OP is confused though, I doubt the card is running with +250mhz offset on the core. Only cards I know that achieve those numbers is the EVGA Kingpin with a 500w BIOS and others at that level.

He's probably just refering to the clock he sees in game, in which case 1900mhz is perfectly normal for a cheap 2080ti if your temp is only 40c. Chances are to get it stable he needs to drop the the boost to under 1650mhz. Just RMA the card.

It can be that 1650mhz is the minimum allowed in the BIOS which is why he maybe can't reduce it.

Again, just RMA the card.

Ah yeah you are right 1650 - I googled just pro and not pro OC. I know he's referring to the clock he sees in game! My 1080ti founders edition out of the box boosted all the way upto 1875Mhz out of the box. You don't have to set an overclock for it to exceed the factory boost clock. It's just the minimum rated. I never said it's running 250Mhz on the core or base clock. If you knock 100Mhz off the base clock it'll still boost to over 1650Mhz anyway. Just not to 1900Mhz like he was seeing.

He cant just RMA the card as he bought it as a whole system from OCUK and they built the machine. JUST RMA it = top diagnosis skillz .

 

[Grim5]

Ah yeah you are right 1650 - I googled just pro and not pro OC. I know he's referring to the clock he sees in game! My 1080ti founders edition out of the box boosted all the way upto 1875Mhz out of the box. You don't have to set an overclock for it to exceed the factory boost clock. It's just the minimum rated. I never said it's running 250Mhz on the core or base clock. If you knock 100Mhz off the base clock it'll still boost to over 1650Mhz anyway. Just not to 1900Mhz like he was seeing.

He cant just RMA the card as he bought it as a whole system from OCUK and they built the machine. JUST RMA it = top diagnosis skillz .

not much diagnostics to be done when the card is already at the minimum clock speed, OP said himself the card has not been overclocked.

 

[Rroff]

The GPU driver is one of the most realtime sensitive parts of the system so a crash in it doesn't necessarily mean it is a GPU problem. If the system is hard locking without graphical corruption then it could be the CPU isn't stable. Testing the GPU in another system if possible as suggested would help to narrow that down I'd also put the RAM through Memtest to make sure it is stable as well before trying anything else.

 

[Grim5]

OP should check his event viewer as well. When my 2080ti started dying, not only did I have crashes like this, but there were many times the driver would crash and actual recover. When this happens the screen will flicker twice and then a message pops up saying "display is connected". Check event viewer logs

 

[Woodsta888]

not much diagnostics to be done when the card is already at the minimum clock speed, OP said himself the card has not been overclocked.

How do you know it's at minimum clock speed? The OP said HE hasn't overclocked it, but OCUK may have, otherwise why would they install Precision X1 by EVGA on his machine with a Palit card? Can't be much diagnosis to do if you make all the wrong assumptions.

Winner!

 

[laserom]

Yeah knock off 100Mhz the core and the memory. Try it and if it stays stable. The factory boost for that card is 1575Mhz and it'll probably on it's own boost higher like you are seeing. It just cant hold it anymore. Wonder if OCUK applied an overclock to the GPU, fine during stability testing but wasn't conservative enough and now just can't hold it?

You should be able to access it through precision X1 unless OCUK set it up to lock their OC or stable (at the time) system. Don't bother trying to run afterburner if precision X1 is being used. Precision X1 is EVGA OC'ing software so maybe OCUK use that on all cards as the techs prefer it. Or the one that built your machine does. 40 degrees is very cool for an aircooled card if it's running at max. Would be in the 70+ degrees plus range if running at 100%. What resolution you gaming at?

I meant it was never more than 40 at time of each crash, but it has been running about 70c at maximum speed but didn't crash at those times I've been gaming. I'm just baffled that it's okay when it runs max speed on 4k monitor but falls over when playing something like Caveblazers at maybe 1400MHZ and 35c, or just browsing the internet. All the dumps seem to point towards Nvidia (driver or hardware).

Actually the stock boost clock is 1650 MHz and only has a power limit of 260w.
I suspect OP is confused though, I doubt the card is running with +250mhz offset on the core. Only cards I know that achieve those numbers is the EVGA Kingpin with a 500w BIOS and others at that level.

He's probably just refering to the clock he sees in game, in which case 1900mhz is perfectly normal for a cheap 2080ti if your temp is only 40c. Chances are to get it stable he needs to drop the the boost to under 1650mhz. Just RMA the card.

It can be that 1650mhz is the minimum allowed in the BIOS which is why he maybe can't reduce it.

Again, just RMA the card.

Used X1 scan, max 2020MHz at 74c.
I meant 40c at maybe 1000Mhz-1400Mhz at time of each crash. see my post above this.

not much diagnostics to be done when the card is already at the minimum clock speed, OP said himself the card has not been overclocked.

I said not manually overclocked, it is factory overclocked.

The GPU driver is one of the most realtime sensitive parts of the system so a crash in it doesn't necessarily mean it is a GPU problem. If the system is hard locking without graphical corruption then it could be the CPU isn't stable. Testing the GPU in another system if possible as suggested would help to narrow that down I'd also put the RAM through Memtest to make sure it is stable as well before trying anything else.

Memtest is good, with 2X8GB Team Force that was in the PC and then replaced and tested the 2x16gb Crucial Ballistix LT from anther machine.

 

[Woodsta888]

Hi Laserom,

I know you said that precision X1 was locked and you couldn't change it. Can you post a pic or tell us what the settings are. I wonder if OCUK have OC'd it. looking at the profile they may have saved in Precision X1 we may see their OC applied.

If it was left at stock, I don't think they would have installed Precision X1 as it's an overclocking utility.

 

[laserom]

OP should check his event viewer as well. When my 2080ti started dying, not only did I have crashes like this, but there were many times the driver would crash and actual recover. When this happens the screen will flicker twice and then a message pops up saying "display is connected". Check event viewer logs

Event Viewer, 7 Criticals relating to Kernel Power (sudden unclean shutdown) for the last month, each crash I've had said the same thing.
What should I be looking for in the other categories (Error, Warning etc)?

Hi Laserom,

I know you said that precision X1 was locked and you couldn't change it. Can you post a pic or tell us what the settings are. I wonder if OCUK have OC'd it. looking at the profile they may have saved in Precision X1 we may see their OC applied.

If it was left at stock, I don't think they would have installed Precision X1 as it's an overclocking utility.

They didn't install it, so there's no profile. I've installed X1 to see the speed and temps not OC (X1 because I've used it in the past as monitoring device on my 2nd screen). Everything is as it was at time of build. All stock. Palit 2080Ti Pro OC is factory OC. Did a X1 scan last night to see if it would crash, reached 2020mhz at 74c, all fine.

How do you know it's at minimum clock speed? The OP said HE hasn't overclocked it, but OCUK may have, otherwise why would they install Precision X1 by EVGA on his machine with a Palit card? Can't be much diagnosis to do if you make all the wrong assumptions.

Winner!

They didn't, no OC was applied by them, I installed X1 as a monitor for temps and speed, no manual OC was applied, it is factory OC.

I apologise for all the misunderstandings above. I've had PCs since 1995, I've had many problems but I've managed to track down all the causes.
But this one is really annoying me with its non-specific issue.
Is it definitely the graphics card whether it be driver or hardware? I just don't know anymore... *cries*

 

[Kelt]

I apologise for all the misunderstandings above. I've had PCs since 1995, I've had many problems but I've managed to track down all the causes.
But this one is really annoying me with its non-specific issue.
Is it definitely the graphics card whether it be driver or hardware? I just don't know anymore... *cries*

There's a new Nvidia driver out today, worth a bash.

 

[laserom]

There's a new Nvidia driver out today, worth a bash.

Cheers, downloading now.
Now, where's that "oh boy here we go again" gif...?

 

[Woodsta888]

I apologise for all the misunderstandings above. I've had PCs since 1995, I've had many problems but I've managed to track down all the causes.
But this one is really annoying me with its non-specific issue.
Is it definitely the graphics card whether it be driver or hardware? I just don't know anymore... *cries*

No need to apologize fella. Just help you get a stable machine for the time being and something you can have a discussion with OCUK if and when you need to call them.


I've just installed Precision X1 myself.

So under the memory and GPU in the 'clock' boxes on the left they read '0'? If yes then the card is at stock.

Slide the memory by 100Mhz to the left so it show -100Mhz
and also for the clock for the GPU so it reads -100Mhz
Click number 1 on the right hand side and then save, bottom right.
Under 'setup' top right hand corner
'General' and tick 'start with OS' and 'start minimized'

Now try and have a game etc. Under the HWM tab after a gaming session note what the boost clocks are doing - will probably still exceed 1650Mhz.

Let us know how you get on.

My machine under certain games will be stable but there are some games like insurgency Sandstorm that are really heavy on the GPU and I got that same crash and it was annoying as it was OK in other games so was hard to pinpoint. Give it a go with the underclock and see if you can get the same crash or it becomes stable.