[SOLVED] GTX 1060 nvlddmkm error 13/BSOD

[W4FFLE]

I recently upgraded my wife's graphics card from an EVGA GTX 760 SC to an EVGA GTX 1060 SC 6gb (06g-p4-6163-kr) and have been having issues with it freezing/BSOD. She plays Ark:Survival Evolved almost exlusively and her GTX 760 just wasn't enough. After a couple hours of gameplay, her game would freeze and her computer would lock up. I checked the event viewer and found 4 instances of the nvlddmkm error ID 13 followed by display error ID 4101 and finally eventlog ID 6008. I also noticed I started getting EnhancedStorage-ehstortcgdrv error ID 10 (category:driver) after the upgrade as well.

I followed steps mentioned in https://forums.evga.com/Basic-troubleshooting-steps-Windows-clean-boot-clean-driver-install-debug-mode-etc-m2504606.aspx and heres what I have experienced..

I was unable properly test clean boot so far due to steam services not being enabled and unable to play steam games.

I ran DDU.exe in safe mode after uninstalling afterburner/PX (and removing saved settings) to delete any existing drivers. After this step I noticed I am now getting BSOD with watch_dog and bad_pool_caller errors. These BSOD also cause the same errors in the event viewer as above.

I have not tried debug mode yet, but I do currently have it underclocked -400 core/-502 mem, which seems to be working for now (more testing needed to be sure).

Things to Note...

I also replaced stock cpu fan/heatsink with Hyper 212 evo and installed an aditional case fan (intake) at the time of the upgrade.
Nvidia Control panel settings - "prefer maximum performance" and vsync is "on" all esle is default.
Afterburner settings default except fan curve set to "auto", default fan curve. (currently also underclocked -400 core/ -502 mem)
I have not tried to overclock this card, just installed drivers and setup nvidia and afterburner/PX (tried both) as above.
Afterburner reported core clock speeds have reached a max of up to 2100mhz while in game prior to underclock (seems high imo).
Unigen Heaven and Furmark don't seem to boost clock card as high as in-game (trying to replicate issue faster)
Max in-game GPU temp is 61c
Max GPU temp during Unigen/Furmark stress test is 67c
Max CPU temp is 63c

Specs are as follows...

Windows 10 64-bit Home
3770k (stock clock speeds)
Hyper 212 evo CPU cooler
16gb (4x4) Ripjaws X 1600mhz
Asus P8-Z77-M motherboard
MX300 275gb SSD
2TB Firecuda
Rosewell Capstone 550W gold PSU
Cosair 350D case (upgraded high air flow front cover) with 2x AF140 (previously 1x) intake, 1x AF120 exhaust fans

Sorry for the long post, just trying to be as informative as possible. Any help/advise would be helpful. I've had an incredible amount of bad luck with defective parts and windows 10 issues recently and my hair is starting to fall out.


EDIT: Update - Issue still persist even while underclocked. Have not tried running this card in my machine yet. Will not be able to try that until Monday.
post edited by Xotic84 - 4 days ago

I have posted this on the EVGA forums and have had no luck so far. Here is a link to the post
https://forums.evga.com/GTX-1060-6gb-nvlddmkm-error-ID-13BSOD-m2879574.aspx

 

[johnbl]

assuming nvidia does not use special error codes, then error 13 would mean invalid data.
this can happen with a stack overflow.
you might go into bios and toggle any setting and save to force the bios to rescan your hardware.
then boot into windows and go to the motherboard vendors website and update the motherboards sound driver.
(hopefully you can get one dated after aug of 2017 because of a common bug that crashes the gpu sound driver)

if you don't use the motherboard sound, you can turn it off in bios or disable it in windows control panel device manager.

if you can not figure out the problem, you might google how to force a memory dump from the keyboard, make the registry settings and then force a memory dump. then copy the files from c:\windows\minidump directory to a cloud server, share the files for public access and post a link.

there are common problems that can cause this issues and the memory dump can point in the correct direction.

you can also download and run autoruns from
https://docs.microsoft.com/en-us/sysinternals/downloads/autoruns
and take a look at the drivers, you would be looking to overclocking drivers or duplicate copies of overclocking driver.

find the menu item to hide microsoft entries, it will reduce the number of entries you have to look at.
(look for 3rd party, old drivers. this website can help identify what the drivers are for:
https://carrona.org/drivers/driver.php?id=atikmdag.sys )

the memory dump can also be scanned for malware corrupted drivers from inside the windows debugger.

I recently upgraded my wife's graphics card from an EVGA GTX 760 SC to an EVGA GTX 1060 SC 6gb (06g-p4-6163-kr) and have been having issues with it freezing/BSOD. She plays Ark:Survival Evolved almost exlusively and her GTX 760 just wasn't enough. After a couple hours of gameplay, her game would freeze and her computer would lock up. I checked the event viewer and found 4 instances of the nvlddmkm error ID 13 followed by display error ID 4101 and finally eventlog ID 6008. I also noticed I started getting EnhancedStorage-ehstortcgdrv error ID 10 (category:driver) after the upgrade as well.

I followed steps mentioned in https://forums.evga.com/Basic-troubleshooting-steps-Windows-clean-boot-clean-driver-install-debug-mode-etc-m2504606.aspx and heres what I have experienced..

I was unable properly test clean boot so far due to steam services not being enabled and unable to play steam games.

I ran DDU.exe in safe mode after uninstalling afterburner/PX (and removing saved settings) to delete any existing drivers. After this step I noticed I am now getting BSOD with watch_dog and bad_pool_caller errors. These BSOD also cause the same errors in the event viewer as above.

I have not tried debug mode yet, but I do currently have it underclocked -400 core/-502 mem, which seems to be working for now (more testing needed to be sure).

Things to Note...

I also replaced stock cpu fan/heatsink with Hyper 212 evo and installed an aditional case fan (intake) at the time of the upgrade.
Nvidia Control panel settings - "prefer maximum performance" and vsync is "on" all esle is default.
Afterburner settings default except fan curve set to "auto", default fan curve. (currently also underclocked -400 core/ -502 mem)
I have not tried to overclock this card, just installed drivers and setup nvidia and afterburner/PX (tried both) as above.
Afterburner reported core clock speeds have reached a max of up to 2100mhz while in game prior to underclock (seems high imo).
Unigen Heaven and Furmark don't seem to boost clock card as high as in-game (trying to replicate issue faster)
Max in-game GPU temp is 61c
Max GPU temp during Unigen/Furmark stress test is 67c
Max CPU temp is 63c

Specs are as follows...

Windows 10 64-bit Home
3770k (stock clock speeds)
Hyper 212 evo CPU cooler
16gb (4x4) Ripjaws X 1600mhz
Asus P8-Z77-M motherboard
MX300 275gb SSD
2TB Firecuda
Rosewell Capstone 550W gold PSU
Cosair 350D case (upgraded high air flow front cover) with 2x AF140 (previously 1x) intake, 1x AF120 exhaust fans

Sorry for the long post, just trying to be as informative as possible. Any help/advise would be helpful. I've had an incredible amount of bad luck with defective parts and windows 10 issues recently and my hair is starting to fall out.


EDIT: Update - Issue still persist even while underclocked. Have not tried running this card in my machine yet. Will not be able to try that until Monday.
post edited by Xotic84 - 4 days ago

I have posted this on the EVGA forums and have had no luck so far. Here is a link to the post
https://forums.evga.com/GTX-1060-6gb-nvlddmkm-error-ID-13BSOD-m2879574.aspx