[Win10Pro x64] BugCheck 0x116 VIDEO_TDR_FAILURE

T

Typhon

Member
Joined
Mar 3, 2017
Posts
20
Location
Athens
Hi all! I need help with my pc as it seems i can't solve that very annoying problem. All happened after a graphics card upgrade from a Radeon HD 5750 to a EVGA GTX 1060 3GB SC Gaming. My problems are related to sudden hardware errors recorded randomly in reliability center and of course at waking up from sleep state. I suspected that the graphics card was defective and i got it replaced. That di not solved the problem. I tried to get some help also from EVGA forums and after a good effort to help me again problem persists. I attach SysnativeFileCollectionApp archive as permon /report runs endllesly in my machine so i can't save the report. If you have any alternative to fix perfmon i'll send you the report asap. Any other necessary information you might need let me know. Thank you.
 

Attachments

Re: BSOD in Win10 Pro

I have driver verifier running, i checked ssd with ssdlife (screenshot) and a question:can i run Windows Memory Diagnostic program? Is it the same with memtest86+? as i see the last is from 2013.
 

Attachments

  • SSDLife.png
    SSDLife.png
    49.6 KB · Views: 4
Re: BSOD in Win10 Pro

Hi Typhon. :welcome:

Did you clean the interior of your machine (pc)?
Try to uninstall completely the popcorn time application and to disable window 10 fast startup (Control Panel\Hardware and Sound\Power Options\System Settings - change settings that are currently unavailable - untick turn on fast startup - click ok. ).
 
Last edited:
Re: BSOD in Win10 Pro

Hi xilolee :wave:
It's a clean new windows installation. I use windows cleanmgr usually, and other cleaners. Didn't help. I have also try disabling hibernation and fast startup by using powercfg /h off. Didn't help. As for the popcorn time application, after reformating my pc i had already many bsods on waking/sleeping states before i even installed it. So maybe that's not the issue.
 
Re: BSOD in Win10 Pro

xilolee said:
I intended the dust inside your machine.
Click to expand...

Cleaned recently with air pressure :grin1:

If a VERIFIER_ENABLED BSOD occurs, please get the dump file from c:\windows\minidump, zip it up & attach to your next post.
Click to expand...

I just had one more crash after waking up from sleep state with verifier.exe enabled. :banghead: Whole /Minidump folder attached.
 

Attachments

Re: BSOD in Win10 Pro

I found on internet a guy that told he re-arranged its cables (in a better way) inside its machine (and this solved his problem), but at the moment I'm not able to find again the thread.
Therefore you could try to reseat all your hardware components (and check if there isn't dust in the slots).

Also try to run sfc /scannow, from an elevated command prompt, without closing it (it could last from 10 to 70 minutes), and report here its final result.
 
Re: BSOD in Win10 Pro

Hi xilolee. I assume that you have examined the attached system files by telling me about the cables. Did you find any errors or other related problems? I have to mention that my rig case is an htpc one that is opened quite often for various reasons including cleaning. My cables are in the best place that they can occupy in a not that big case (DH101). The only hardware that sits are the ram modules and the vga. All are tight. The command sfc /scannow reported that all files are ok.
 
Re: BSOD in Win10 Pro

I examined all the logs but I'm not a BSOD analyst, hence I could have missed something.
It could be a good idea to try memtest86 (without the +) because it seems that's the original software memtest86+ derived from and also because it seems it's the more updated at the moment.
I don't know if running the windows memory diagnostic could be the same, but I know all windows users can modify its settings to be more scrupulous, especially windows 10 users (I have limited experience with vista and seven, but I think their users can also modify the test).


Edit:

I tried the windows memory diagnostic tool: test mix extended, cache on, pass count 5.
It took about 2H38min, i.e. an average of about 31 minutes per test.
Therefore I can estimate the total amount of hours (on my machine) could be about 8, if I had chosen 15 passes.
Using Pentium G645 and 6GB RAM; from piriform speccy:
  • Nanya Technology NT4GC64B8HG0NF-CG DDR3 4096MB PC3-10700 (667 MHz)
  • Nanya Technology NT2GC64B88G0NF-CG DDR3 2048MB PC3-10700 (667 MHz)
I found two events (results) in event viewer - windows logs - system:
Read More:


How to enable the windows memory diagnostic tool in windows 10
(although this could be valid also for previous windows versions, from windows vista onwards)

  1. (Classic shell disabled) Click windows start, windows administrative tools, windows memory diagnostic (and you can double click it).
    or
    (Classic shell enabled - xp style) Click windows start, Windows, programs, windows administrative tools, windows memory diagnostic (and you can double click it).
    or
    Click windows start, type directly memory and the windows memory diagnostic app should be the first hit (and you can double click it).
    or
    Go to C:\WINDOWS\system32, find and double click mdsched.exe.
    or
    Right-click windows start, click run, copy/paste mdsched.exe in the run window and press enter.
  2. Click yes to the UAC window that should appear.
  3. The windows memory diagnostic tool will ask you either to restart now or to schedule the test at next windows startup.
    (I didn't use the exact same words, but the sense is the same).
  4. You can change the options of the tool pressing F1 (key on your keyboard).
  5. The first option you can change is the test mix:
    • basic --> MATS+, INVC, SCHCKR (cache enabled)
    • standard --> like above and LRAND, STRIDE6 (cache enabled), CHCKR3, WMATS+, WINVC
      Complete list: MATS+, INVC, SCHCKR (cache enabled), LRAND, STRIDE6 (cache enabled), CHCKR3, WMATS+, WINVC
    • extended --> like above and WINVC, MATS+ (cache disabled), STRIDE38, WCHCKR3, ERAND, STRIDE6 (cache disabled), CHCKR8
      Complete list: MATS+, INVC, SCHCKR (cache enabled), LRAND, STRIDE6 (cache enabled), CHCKR3, WMATS+, WINVC, MATS+ (cache disabled), STRIDE38, WCHCKR3, ERAND, STRIDE6 (cache disabled), CHCKR8
  6. (You can go to the other options using the TAB key on your keyboard.)
    The second option you can change is the cache: default, on, off.
  7. The third option you can change is the pass count, from 0 (zero, nought) to 15.
    0 (nought) means 15.
  8. Once done, press F10 (key on your keyboard) to apply the changes.
  9. The test will start automatically after the changes.
 
Last edited:
Re: BSOD in Win10 Pro

Ran memtest86 for 4 hours and no errors detected. Seems RAM is fine. Windows memory diagnostic reported no errors also.
 
Re: BSOD in Win10 Pro

I uninstalled popcorn time completely, corsair link and disabled fast startup as you suggested. Tonight turning on the system i had one more bsod & restart. That's really strange. Have a look at this: I have a Logitech keyboard K800 which uses a tiny usb unifying receiver (wireless). Logitech offers Setpoint as suplementary software for the device. It's pretty useless as all functions can be access from the keyboard itself. Not using it. But it installs a driver. After installing it again yesterday i noticed a new device showed up in Device Manager in Keyboards AND the pc sleeps and wakes 99% without crashing. I had only one crash that happened with verifier enabled on wake up. My problem continues when starting up the pc after shut down. :huh: :huh: Anyway i'll test the system more and see what happens.
 
From the last 06-03 bsod i'm good until today. To narrow down a bit i replaced logitech unifying usb receiver with a new one (keyboard also but i do not think it's relevant), installed the logitech setpoint software, unticked "fast startup" option AND uninstalled completely corsair link & popcorn time apps. My opinion is that microsoft maybe installs a generic driver for logitech unifying receiver that is incompatible/conflicting with mobos power managemet. An other very rare chance is that my evga graphics card has some kind of incompatibility with one of the uninstalled programs. Just to know, what made you suggest the uninstallation of those programs?
 
0x116 bugcheck in your latest dump - so still showing issues with the GPU or drivers. Nvidia drivers & DirectX are showing up everywhere.

Code: 
ffff8980`3eb55b98  fffff80a`437bba88 [B]dxgkrnl[/B]!TdrBugcheckOnTimeout+0xec
ffff8980`3eb55ba0  00000000`00000116
ffff8980`3eb55ba8  ffffd380`31d6d010
ffff8980`3eb55bb0  fffff80a`45e827b0 [B]nvlddmkm[/B]+0x9a27b0
ffff8980`3eb55bb8  ffffffff`c000009a
ffff8980`3eb55bc0  00000000`00000004
ffff8980`3eb55bc8  00000000`00000000
ffff8980`3eb55bd0  ffffd380`31d6d010
ffff8980`3eb55bd8  fffff80a`4379e92f [B]dxgkrnl[/B]!ADAPTER_RENDER::Reset+0x153
ffff8980`3eb55be0  fffff80a`45e827b0 [B]nvlddmkm[/B]+0x9a27b0
ffff8980`3eb55be8  ffffd38f`e9cbe690
ffff8980`3eb55bf0  ffffd38f`e9b76010
ffff8980`3eb55bf8  ffffd38f`e9cbe718
ffff8980`3eb55c00  ffffd38f`eb637800
ffff8980`3eb55c08  fffff80a`4379badf [B]dxgkrnl[/B]!DXGADAPTER::Reset+0x307

NVidia driver looks up to date too:
Code: 
fffff80a`454e0000 fffff80a`46308000   nvlddmkm T (no symbols)           
    Loaded symbol image file: nvlddmkm.sys
    Image path: nvlddmkm.sys
    Image name: nvlddmkm.sys
    Browse all global symbols  functions  data
    Timestamp:        Thu Feb  9 22:20:40 2017 (589CEB38)
    CheckSum:         00DDCA08
    ImageSize:        00E28000
    Translations:     0000.04b0 0000.04e4 0409.04b0 0409.04e4

If you've already replaced the card, and already reinstalled Windows, then I have to assume this is some other hardware issue.

Wondering if this is due to a failing/underpowered PSU? What PSU do you have? The HD5750 draws 87W at load, the 1060 draws 120W. Not much more, but could be enough.

Any BIOS updates for the machine?

Memtest86+ should also ideally run for 8+ passes, but I doubt memory is the main cause here.

-Stephen
 
Hi Tekno Venus, :smile9:
My pc is very old and asus have ceased support long ago, though i have the latest bios which is from 2010. The vga is replaced with a brand new 10 days ago. My psu is a Corsair RM series 750W 80+ Gold still in warranty.

As for the Nvidia driver, i have already wipe every trace in safe mode with DDU and clean reinstalled. It's not the first time and it never really helped.
Would you prefer to use memtest86+ instead of memtest?

P.S. Click my specs. Everything is filled...
 
Last edited:
Hi,

Well, that PSU is certainly good enough then! Nice choice - I have a smaller Corsair SF unit in my rig and am very pleased with it!

You seem to have pretty much tried everything I would suggest as a starting point. Do you have another computer you could try the GPU in? Since these BSODs only started when you put the new GPU in, they must be related to that somehow. I wonder, if however unlikely, you have got a second dead card... :(

Here is some more information on 0x116 BSODs if you're interested in some reading: https://www.sysnative.com/forums/bs...rmation/35-video-tdr-timeout-0x116-0x117.html

Am only on my laptop at Uni at the moment, so don't have access to the debugging tools right now.

Yeah, Memtest86+ is what I would recommend. Our tutorial is here: https://www.sysnative.com/forums/hardware-tutorials/3909-test-ram-memtest86.html. It really needs >8 passes, which can take a long time so best to leave it overnight.

-Stephen
 

Has Sysnative Forums helped you? Please consider donating to help us support the site!

Back
Top