New computer build get bluescreen and crash typically during gaming

Hi

I could really use some help. Tried everything I can think of to find the problem. This is what I have tested so far

Ran memtest86+, crashed after running for an hour, but on the second try it ran for ours and I got no errors
also tried a tool called tm5 and it found no errors

Then I tried to take out all memory cards except one and ran FurMark for a few hours, it seems to work fine, but then while gaming GTA 5, the PC crashed again. So not sure if it's something wrong with the memory cards.

I tried to change nvidia settings:

disable v-sync
Tried to set Power Management Mode to Maximum Performance.
Tried different versions of drivers
update sound drivers
Installed latest bios
disabled intel virtual machine in bios...

Tried to run these powershell commands:

Code:

Dism /Online /Cleanup-Image /CheckHealth
Dism /Online /Cleanup-Image /ScanHealth
Dism /Online /Cleanup-Image /RestoreHealth

sfc /scannow (here it said it found some issues that where fixed)

Uploaded all the log files, it does say

Code:

dxgkrnl.sys	dxgkrnl.sys+26665e	fffff803`50450000	fffff803`507fa000	0x003aa000	0x78205bd5	12.11.2033 07:36:21

In the Minidump, these seems to indicate problems with the memory or graphic card, but the graphic card worked just fine in my previous build so I think the card is fine.

EDIT: Just ran a tool from OCBASE/OCCT : Free, all-in-one stability, stress test, benchmark and monitoring tool for your PC, when I ran the VRAM stress-test, my computer instantly crashed. Is my graphic card broken?

Windows version number: 21H1 (OS Build 19043.1266)

It could be the card, but the first step is if you have set the ram at 1.35 voltage to use it at tested speed 3600
https://www.corsair.com/uk/en/Catego...tab-tech-specs

set it 1.2 to use SPD at 2666
or whatever the BIOS settings on that board require to run SPD on ram - serial presence detect

I will look at the logs , but try this step first

EDIT From your computer spec you say
Corsair Vengeance Black RGB LED Pro DDR4 3600MHz 2x8GB

so I thought you had two sticks of ram to make 16GB
In fact you have four sticks as I am sure you know
From your logs - you have all four DIMMS populated

Your OS was installed on 30 Sept
has the system crashed since that date or is it a more recent event

VERA Crypt may be the cause if when checking ram settings it is at 1.2

There does seem to be a suggestion in one of the dumps that the graphics may be the problem
0x116_IMAGE_nvlddmkm.sys
VIDEO_TDR_FAILURE (116)
which in simple terms means
The VIDEO_TDR_FAILURE bug check has a value of 0x00000116. This indicates that an attempt to reset the display driver and recover from a timeout failed.

However that does NOT mean that it is the card or indeed the driver.
I need to exmaine more dumps and logs I will do so when you reply

=============
In the meantime if you wish use the onboard graphics in the processor

Intel® UHD Graphics 750
HDMI port on I/O plate
disconnect card and either take it out or set BIOS to use onboard only
That will eliminate any question of the card or its driver.

Thank you so much for your reply, I think I am on another timezone so might take a while before you get a response.

I have never manually changed voltage in bios before, after checking it was set to "auto" and it looked like it was set to the correct values by default.

I manually changed it to the values as seen here:
Imgur: The magic of the Internet

Your OS was installed on 30 Sept
has the system crashed since that date or is it a more recent event

It has happened every day since I built the computer About 30min to an hour into a video meeting or while gaming.

I do run veracrypt, I have encryped the entire system drive. I have also three more 4 TB drives that are encrypted with veracrypt and mounted when I start windows.

From your logs - you have all four DIMMS populated

Yes that is correct, was a bit quick to copy paste my specs. Updated it now.

Intel® UHD Graphics 750
HDMI port on I/O plate
disconnect card and either take it out or set BIOS to use onboard only
That will eliminate any question of the card or its driver.

I can try this.

For now I have changed the voltage and RAM speed manually in BIOS. Then I ran DDU v18.0.4.4 to remove all NVidia drives and installed the latest version 472.12-desktop-win10-win11-64bit-international-dch-whql. Will run furmark for a while, and see if it crashes.

There is no need to run DDU and in fact you should not do so, in the normal course of events, it is designed for use when there are problems
As DDU forum says
Hi, using DDU is not mandatory. It's best to use it when you need it. If you want to keep your profile informations, use the normal nvidia installer. If you encounter weird issue, use DDU.
Is DDU worth using before installing new drivers? - Wagnardsoft Forum

you should use the recomended Nvidia way - settings - apps which is of course control panel programs and features
It is only IF there are difficulties experienced that it is recommended to use DDU

Let me know please how it goes, my money is on VeraCrypt, if the ram setup does not solve it.

Maybe worth mentioning, the displayport cabel used to connect my graphic card tend to get stuck, so I have been pulling and twisting quite hard trying to get the cabel out. Could this have broken the card?

Unlikely - as of course the connection is secured on the plate of the card is it not - but possible
more likely IMHO is the cable connection itself
Try another cable

Pleased to have helped a little
Let me know how you go on
It may be worth if a new DP cable does not work trying another cable to a another port on the card if the monitor allows this
Just in case the twisting has damaged a pin/socket on that port

Have you tried the other DP depending on the manufacturer of the card I think it has 2 DP and 2 HDMI
so - I am not suggesting to buy an HDMI lead but if you can borrow one and the monitor has HDMI you could try that