[SOLVED] Games crash to desktop with no errors. Out of ideas

[Varg7]

I recently built a brand new PC (in the last month) and no matter what game I play, at some point it ends up crashing to the desktop with no errors. Never had this issue with previous computers either. Thing is sometimes it happens within 10 minutes of playing, sometimes it doesn't happen for hours. I keep a close eye on the temps and they all seem fine, so I don't -think- it's that.

Things I have tried so far
-Full GFX driver reinstalls (using DDU)
-Updated windows as much as possible
-Memtest86
-Disabled XMP
-Moved to Windows 11, then back to Windows 10 a handful of times
-Double checked all hardware connections
-Tried previous GFX drivers
-Updated BIOS (from F8 to F20 to F21)

Seems at this point it's likely a hardware issue perhaps? Even though everything was bought brand new. The main culprit I suspect could be the GPU. A few issues with it aside from the crashes such as shadows/lighting/random flickers during gaming, terrible coil whine during gaming and also noticeable coil whine on the desktop when just moving the mouse. I also had a couple of bluescreens (Kernel error 141) and hard crashes where the entire screen goes all pixelated and just freezes, forcing a hard reset. But those were very rare. I just get the feeling the graphical side of the PC just isn't totally stable.

I'm no PC guru so if anyone can think of anything else that might be the cause it would be very much appreciated. Spent a fair amount of money on this PC and not really got much enjoyment out of it since I bought it because I use it just waiting for something to go wrong. And it inevitably does.

Note: When I run things like Unigine Benchmarks, it occasionally stutters and there is the odd shadow flicker here and there, but it has never crashed. The issues I face are only when gaming.

Specs
-ASUS TUF 3080ti OC (GPU)
-i7 12700k (CPU)
-Gigabyte Z690 Gaming X (Mobo)
-Corsair Vengeance DDR5 5600 MHz (RAM)
-Firecuda 530 1tb (SSD)
-Corsair RM1000e (PSU)

Thanks!

 

[Colif]

does anything show in reliability history?

Does anything show on the follow:
open system information app
in the left hand column
tick the + next to software environment
open Windows error reporting
Depending when the last install was, it might take time to open

don't click on header for dates in right column as it doesn't sort them right, it uses alphanumeric meaning it will sort by 1st day, not actual month.
top section shows application crashes, section below it is windows errors.
I am curious if you getting BEX or Pre Radar errors.

just an example of what you might see

in case you get more BSOD
Can you follow option one on the following link - here - and then do this step below: Small memory dumps - Have Windows Create a Small Memory Dump (Minidump) on BSOD -
that creates a file in c windows/minidump after the next BSOD

1. Open Windows File Explore
2. Navigate to C:\Windows\Minidump
3. Copy the mini-dump files out onto your Desktop
4. Do not use Winzip, use the built in facility in Windows
5. Select those files on your Desktop, right click them and choose 'Send to' - Compressed (zipped) folder
6. Upload the zip file to the Cloud (OneDrive, DropBox . . . etc.)
7. Then post a link here to the zip file, so we can take a look for you . . .

 

[Varg7]

Hey thanks for the reply. So I did what you said and below is what was there. I see at least one of each that says:

• RADAR_PRE_LEAK (02/12/2022 18:49) - this one is something to do with the game Destiny 2.exe when I scroll the slider further to the right.

• BEX (03/11/2022 23:27) - this one is mentioning something about OneDriveSetup.exe when I scroll the slider further to the right.
  
  Probably worth mentioning it happened again earlier while I was playind Red Dead Redemption 2, which I play through Epic Games Launcher. So I assume that the latest one at the top is something to do with that.

I also did the steps for if I get another BSOD but I haven't had one of them for over a week now so not sure how long it might take to get that info.

 

[Colif]

Try running Seatools on your nvme - https://www.seagate.com/au/en/support/downloads/seatools/

A BEX error is a “Buffer Overflow Exception”. This error is typically raised when Microsoft Windows detects that a program tries to put more data than possible in a region of memory. This is the computer equivalent of overfilling a glass with water until the water spills over.
its an application error, in that its a problem with the program. Try uninstalling and reinstalling it again.

Radar Pre Leaks are similar, they are memory leaks but in the programs themselves, not windows. Windows identifies the problem and closes program before it gets out of hand.

i hate how first 10 are from December and then it jumps back a month to Nov. I actually complained about it to Microsoft but I don't expect it to change. rest of report is last month... not really useful to us.

 

[Varg7]

Sorry for the later reply, was at work. Okay so I ran the SeaTool and did a handful of the different tests and they all passed.

Radar Pre Leaks are similar, they are memory leaks but in the programs themselves, not windows. Windows identifies the problem and closes program before it gets out of hand.

So with that in mind could that be something to do with the fact I have XMP enabled? Up to 5600MHz from 4800. If I recall, the last time I disabled XMP the issue I came across was different to the random desktop crash. Pretty sure that was when I got the hard crash that required the full restart. But I think that was just a one time issue as I never had that problem again after I used DDU to reinstall all my graphics drivers.

 

[Colif]

no need to rush back, its your post... i should be asleep by now.

try running without xmp and see if it helps

 

[Varg7]

Just seems a shame that if it worked and it turns out that was part of the issue, I would then have to run at less than what was paid for. Say that was the case and it was to do with XMP, could there be a possible work around for that? Perhaps manually setting it all up instead of using one of the automatic profiles?

I was running a game with XMP off then, and of course it hadn't been running long enough to know whether or not it was fully fixed from desktop crashes (as some times it can take hours before it happens) however I didn't notice any of the shadow/lighting flickers I had noticed before.

 

[Colif]

Corsair Vengeance DDR5 5600 MHz (RAM)

what are the codes for the ram? what capacity?

corsair haven't tested any 4800 mhz Corsair sticks, highest they show is 4600
fastest ram Corsair tested with motherboard is 3600, its all ddr4 though? Their website seems broken to me.

 

[Varg7]

It's definitely 4800, it even says that in both the BIOS and Task Manager. Then when I enable XMP it goes up to 5600MHz. Though there seem to be 2 ways you can enable the XMP in the BIOS, one where you just select XMP 1 and another where you have all these options with things like Samsung X.M.P. 5600 36-36-36-76-1.250 and others that are similar just with different names instead of Samsung, such as Hynix & Micron. All for various different speeds.

This is it below right now as stock with XMP off.

Though when I turn XMP on it goes to 5600 (2800.0 DRAM Frequency) and it also changes to 36-36-36-76.

 

[Colif]

It doesn't help there are 2 revisions of the motherboard and one is only DDRr4

list of sticks is suddenly shorter - https://www.gigabyte.com/Motherboard/Z690-GAMING-X-rev-10-11/support#support-memsup
there are 3 there that run at 5600
CMT 32GX5M2X5600C40 ve r 4. 43
CMK32GX5M2B5600C36 ve r 4. 43
CMT 32GX5M2X5600C36 ve r 4. 43

you might have one of those, I can't tell. You can tell by looking on the SPD tab of CPU-Z, it shows it on Part Number.

Corsair only tested the DDR 4 model so I can't go back on them for confirmation that ram works with board.

 

[Varg7]

Yeah looks like I have the correct ram for that. And XMP does work technically as sometimes I can play games for hours with it enabled and have no issues. But my thought is it could also be the reason that things go unstable and crash at times as well? I don't really know much about XMP to be honest. I just enable the profile in the bios and thats about as far as my knowledge on it goes unfortunately.

 

[Colif]

did we disable XMP?

is it in any game or specific ones?

kernel event 141 could be GPU, can be caused by heat so clean inside a case and remove dust from heatsinks.
Are you overclocked? as that can cause it

Did we look in reliability history?

with no actual errors, it makes it difficult. Its not restarting as such, just crashing games

try running this: https://www.intel.com/content/www/us/en/download/15951/19792/intel-processor-diagnostic-tool.html?

i think if it was a driver it would mention it somewhere.

I will try not to say its the GPU.

you could remove GPU and run off the IGPU and see if you have same problems, although gaming might be hard.

 

[Varg7]

-XMP is currently disabled as of yesterday.
-No specific games, it has happened at least once on all of them. Though it seems to be getting much longer before it happens now, which is odd.
-The GPU and PC are brand new so there is no dust at all, the GPU never gets above 66c during gaming. It is not overclocked at all by me, but it is the OC edition, which I assume means it comes with a bit of a factory overclock. That said it's been about 2-3 weeks and I have not seen that Kernel error again yet.
-Yeah checked out reliability history and there is nothing recently in regards to the recent crashes.
-I ran NovaBench last night and on it's first attempt it reset the whole PC just as it got to the GPU testing stage. But then I ran it again twice after that with no issues & the GPU actually performed well above average compared to others the same.
-I ran CineBench for 10 mins with no issues.

I do have the option to RMA the card. Problem is I got it at quite a good price while they had a deal on for it. Now all the cards are back up to high prices and they are all sold out again. So ideally, I'm hoping it's not the card as I would then struggle to get hold of something to replace it with. Unless of course I just wait until the 13th when the new AMD cards are released and see what I could get my hands on.

===================================================================

Update:

try running this: https://www.intel.com/content/www/us/en/download/15951/19792/intel-processor-diagnostic-tool.html?

Passed

 

[Colif]

i wouldn't expect problem to be PSU. I would expect if it was PSU it would turn screen off, not CTD.

If you ran off the igpu and still crashed, it would remove GPU from list.

 

[Varg7]

i wouldn't expect problem to be PSU. I would expect if it was PSU it would turn screen off, not CTD.

If you ran off the igpu and still crashed, it would remove GPU from list.

Would my games even run like that haha? They are pretty graphic intense games. I could try I guess, sure.

 

[Colif]

i don't know. probably not well. the igpu in the intel cpu aren't amazing

only other way is test gpu in another PC.

 

[Varg7]

Okay so today the PC is deciding it's really had enough.

-3 game crashes to desktop within 20 mins, the game (BF:2042) stutters, then freezes, then drops out to the desktop.
-As I was going in to Nvidia Geforce Experience to try to grab a recording I did, I get a BSOD.
-Random black screen flickers every so often while just using the PC on the web etc. This is pretty rare though, just something I noticed a couple of times.

Other people who have posted up about the random desktop crashes have had suggestions given to them about upping the XMP voltage or something. Could this be something to do with it?

Anyway here is the info on it all:

Dropbox BSOD MiniDump
https://www.dropbox.com/s/xd0qeyxopwx521y/120522-6265-01.zip?dl=0

(During the actual Blue Screen there was a big block pixel missing about the size of a pea)

Note: I have yet to remove the GPU. This is all with it still attached.

 

[Varg7]

Also worth mentioning that last night I decided to try a different XMP Profile. I swapped from a profile that was named Samsung X.M.P. 5600 36-36-36-76-1.250, which it is the one it is usually on. Over to one just named XMP 1. Both are exactly the same in terms of numbers though. I think -maybe- the XMP 1 was causing these issues today? as it's never usually that often that the crashes happen. That said, I have no clue of the difference between the two profiles as both are 36-36-36-76 - 1.250.

Samsung X.M.P. 5600 36-36-36-76-1.250 is in a BIOS folder called: DDR5 XMP Booster
XMP 1 is in a BIOS folder called: Extreme Memory Profile (X.M.P.)

I just wanted to give that a quick shot before I started to use the PC with XMP disabled entirely.

 

[Colif]

i have a bit of a headache so my ability to help is reduced, i will have to look at this again later.

conversion of dump

report - click run as fiddle to read


File: 120522-6265-01.dmp (Dec 6 2022 - 06:00:42)
BugCheck: [SYSTEM_SERVICE_EXCEPTION (3B)]
Probably caused by: ntkrnlmp.exe (Process: Video.UI.exe)
Uptime: 0 Day(s), 12 Hour(s), 19 Min(s), and 22 Sec(s)
Process named is victim, not cause

Comment: The overclocking driver "RTCore64.sys" was found on your system. (MSI Afterburner)

seems it all happened in virtual memory. stack pointer error

Corsair AIO -not blaming it, just noticed its drivers
what Asus do you have?
Creative soundcard?

May 27 2020

AmdTools64.sys

AMD Special Tools driver

another intel system with Nvidia drivers that randomly has a AMD driver that talks to bios
link to how to find it - https://forums.tomshardware.com/threads/identify-what-device-a-driver-is-used-by.3785876/

Jun 18 2020

RTCore64.sys

!!! Overclocking Software - RivaTuner - MSI Afterburner http://www.msi.com/ or EVGA Precision X http://www.evga.com/

there is a fake version of that that is really a virus but if you got it off MSI its real.

at least if you use the intel GPU the drivers are already there.
rgb fusion is buggy but I don't think its cause.

 

[Colif]

in a BIOS folder called: DDR5 XMP Booster

AORUS RGB DDR5 6000MHz 32GB memory kit can enable “DDR5 XMP Booster” function which detect the brand of memory IC in the BIOS setting to allow users to choose promptly from multiple built-in memory overclocking profiles, boosting the native DDR5 or XMP DDR5 memory speed.

https://www.gigabyte.com/Press/News/1969

not a lot out there, I would stick to the XMP 1 as its less specific

Could be they still need to release bios updates to further enhance compatibility
its mentioned here:

 

[Varg7]

Yeah it's fine don't worry if you have a headache. I really appreciate the help up to now as I have been struggling with this for a month. Nobody else has helped as much as you have.

-I'm sorry though but I don't really understand what a lot of what you posted in the first reply means. Could you give me perhaps a simpler version haha?
-When you ask which Asus I have, what exactly do you mean?
-I have uninstalled my current Afterburner and will get the one from MSI directly.

Also should I follow the enable of XMP exactly as shown on that video? That video is from a year ago and they released a BIOS update (F21) about 2 weeks ago, which specifically mentioned improvements on DDR5 32gb 5600Mhz compatibility. Which is exactly what I have. I upgraded to that F21 BIOS about a week ago.

Thanks alot

 

[Zerk2012]

Yeah it's fine don't worry if you have a headache. I really appreciate the help up to now as I have been struggling with this for a month. Nobody else has helped as much as you have.

-I'm sorry though but I don't really understand what a lot of what you posted in the first reply means. Could you give me perhaps a simpler version haha?
-When you ask which Asus I have, what exactly do you mean?
-I have uninstalled my current Afterburner and will get the one from MSI directly.

Also should I follow the enable of XMP exactly as shown on that video? That video is from a year ago and they released a BIOS update (F21) about 2 weeks ago, which specifically mentioned improvements on DDR5 32gb 5600Mhz compatibility. Which is exactly what I have. I upgraded to that F21 BIOS about a week ago.

Thanks alot

Enable XMP then manually increase the voltage to 1.3 from 1.25.

If that fails go to the start menu and disable everything that is not needed for the PC to run, no Discord, keyboard software, MSI afterburner, RGB control software disable all of that Everything and try again.

Last time I told somebody to disable everything it stopped the crashes then he started enabling them one by one and ended up the Logitec keyboard software was causing the crashes.

 

[Varg7]

Would I be correct in assuming its VDD & VDDQ that I need to change to 1.30v?

So like this...

OLD

NEW

 

[Varg7]

Putting the voltage up to 1.3 didn't work. Worth trying to go higher, up to 1.35 perhaps?

 

[Colif]

When you ask which Asus I have, what exactly do you mean?

I had a headache, remember... I missed this

ASUS TUF 3080ti OC (GPU)

which answers my questions. Do you have Asus GPU Tweak installed?
You have 2 Asus drivers running, I was just curious why.

May 27 2020

AmdTools64.sys

AMD Special Tools driver

This doesn't belong on your PC.
You have an intel system with Nvidia GPU that randomly has a AMD driver that talks to bios
link to how to find it - https://forums.tomshardware.com/threads/identify-what-device-a-driver-is-used-by.3785876/

The link shows how to find where the driver is running on PC and how to stop it. I don't know how/why its on your PC as you have an Intel system

The bsod itself happened in ram or on the CPU. That really isn't saying anything as that is where most operations on PC happen. I can't tell exactly what cause was.