• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

all games crashing

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

c00l3st

Member
Joined
Oct 10, 2011
most errors are caused by dwm not responding (it crashes and restarts) it tags the gpu as the main source.
i am very frustrated as this has been going on for more than a week.




Things i tried :
all installations are done using clean/full wipe method including the use of DDU for gpu.
Note : Most recent install of windows 10 is from 1803-updated all the way to 2004.

Install win10: 1803,1902,1909,2004

Nvidia driver : 456.55 , 456.38 , 452.06, 451.67, 451.48,446.14. Using 451.67

Updated bios from 2008 to 2202
running bios v1821- will update if this fixes it (was able 2 play 2 games of warzone plunder) -updated bios again to 2202.

unplugged everything and re plugged (with a little more pressure) - seems to work fine as i have ran uni engine heaven benchmark more than 5x now, without crash and errors.
still a few games crashing

sfc /scannow , dism , online dism ,
sfc /scannow(scannow /f) seems to be a bit buggy to me, earlier today i had one instance that it confirmed to have repaired all corrupted files. in the afternoon and this evening it says it cannot repair the corrupted files, and after several tries it went back again to its cycle of repairing it successfully and not.

disabled CSM in bios - after this all the sfc /scannow errors are gone. system is no longer freezing and dwm.exe is no longer crashing.


2700x
16gb kingston 3000mhz
120 250 gb kingston ssd , 250gb nvme lexxar , 1tb hdd
Asus B450 prime plus
rtx 2070 super (brand new) asus strix model
Corsair RM750




any input is highly appreciated.
Thanks !
 
Last edited:

Robert17

Premium Member
Joined
Feb 1, 2011
When updating NVidia drivers, do you completely uninstall the previous driver before installing the new one? This could potentially make a difference. Also, note that installations of newer Winderz releases can roll back drivers for video cards; you have to keep an eye on Redmond.
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
Most recent install of windows 10 is from 1803-updated all the way to 2004.
That and all your corrupted/fixed/corrupted/cant fix symptoms......have you tried a fresh installation of Windows with the latest version already there (from the MS page)?

I'd wipe your SSD and format it fresh and install windows 10 fresh and go from there.
 

BugFreak

Joined
Apr 29, 2010
Location
Central FL
Are you at stock speeds on everything? Have you run any stress tests (both not graphics related and graphics tests) on it to see if it can handle those? Also make sure your ram voltages are proper for your sticks.

Since you just upgraded the GPU I would usually look to a driver problem but you seem to be on top of that. I would also consider the PSU not being able to power it with the new GPU upgrade but a Corsair 750 should be plenty for your setup. Because of all that I would start by resetting my BIOS then checking the voltages to ensure they are properly set. This may not necessarily mean stock voltages but what you now the piece runs stable at.

If all that fails then I would agree with the above statements of a fresh Windows install.
 
OP
C

c00l3st

Member
Joined
Oct 10, 2011
hi guys, thanks for responding!


all installations are done by clean/fresh/full wipe.
all installations of gpu driver are done via DDU.
the last install i did of 1803 earlier , and upgraded it all the way to 2004.
all are at stock speeds and default settings including docp 2666 instead of 3000mhz.

im out of options but to try out older bios.

so far ive only tested 2 bios. 2202 and 2008
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
all installations are done by clean/fresh/full wipe.
I'm confused. To be clear, you have formatted the OS drive and used the latest version of windows (without upgrading) to install, correct?

DOCP isn't default, typically. DOCP enables the profile on the memory. If you set optimized defaults, it should revert to the JEDEC specs on the RAM.
 
OP
C

c00l3st

Member
Joined
Oct 10, 2011
I'm confused. To be clear, you have formatted the OS drive and used the latest version of windows (without upgrading) to install, correct?

DOCP isn't default, typically. DOCP enables the profile on the memory. If you set optimized defaults, it should revert to the JEDEC specs on the RAM.

each install of mentioned version of windows was done and tested. i am currently running 2004 from the fresh install of 1803 that i upgraded today.
i did try each version.
(i tried different versions to check if it fixes it)
ram is not set to docp. its set to auto.
 

The_Jizzler

Member
Joined
Feb 7, 2003
If sfc keeps repairing then failing, try a chkdsk on your drive. Not sure which dism options you've tried but there's a few to use and an order to do it that may or may not be successful. You could always do an over the top install of windows, just tell it to keep your files and settings. Id still run sfc even after that was done to confirm all is well.
 
OP
C

c00l3st

Member
Joined
Oct 10, 2011
hi guys, thanks all again for all your effort. i have updated the status, but some games specially cod warzone is still crashing.

maybe i can cosider this a bit fixed since most the games are working.
Few bugs found on some bios versions lower than 2202:
Pcie x16_1 slot gets stuck on 1.1 2.0 and will take a second before it goes to 3.0 (before it reaches 3.0 the game already crashed)
another weird observation i got from this was on task manager if you see the app usage of a d3d app, it should say d3d-0 or 0-d3d , when i was using the previous bios it always says GPU-(COPY)
will continue to update if i discover something else.


games that doesnt crash anymore - diablo 3 , dota 2, apex legends , ffxiv.
Apps not crashing - uniengine heaven benchmark , furmark , stress test from gpu-z , and benchmark and stress test from cpu-z.

- - - Auto-Merged Double Post - - -

hi guys, thanks all again for all your effort. i have updated the status, but some games specially cod warzone is still crashing.

maybe i can cosider this a bit fixed since most the games are working.
Few bugs found on some bios versions lower than 2202:
Pcie x16_1 slot gets stuck on 1.1 2.0 and will take a second before it goes to 3.0 (before it reaches 3.0 the game already crashed)
another weird observation i got from this was on task manager if you see the app usage of a d3d app, it should say d3d-0 or 0-d3d , when i was using the previous bios it always says GPU-(COPY)
will continue to update if i discover something else.


games that doesnt crash anymore - diablo 3 , dota 2, apex legends , ffxiv.
Apps not crashing - uniengine heaven benchmark , furmark , stress test from gpu-z , and benchmark and stress test from cpu-z.
 

Nebulous

Señor Senior, Senior
Joined
Oct 11, 2002
Location
The Empire State
Me thinks your games drive is failing or memory is failing. Also instead of having memory on "AUTO", set timings/values/voltage manually to their respected specs and run memtest.
 
OP
C

c00l3st

Member
Joined
Oct 10, 2011
do you think it was because i was only using 1 pcie power cable from the psu? now i tried to use 2 of them direct from the psu to the gpu, it seems to be alot better i was able to play for almost 2 hours.
 

Robert17

Premium Member
Joined
Feb 1, 2011
Yeah, it's a pretty good idea to mention these things early on in the post. If a GPU has two connector leads for power, always fill them both.
 

don256us

Uber Folding Senior
Joined
Jul 17, 2003
In reading this post, from the very onset, my default answer is that the power supply is not putting out enough to run 3D mode on your graphics. I see that you have a 750 Watt good name brand. When you then added that you hadn't connected all of the power connectors, that confirmed my suspesion in a way I hadn't anticipated. Going from 2D to 3D takes a lot of power and will reveal a weak power supply.
 
OP
C

c00l3st

Member
Joined
Oct 10, 2011
thanks for inputs ! well yeah the single cable from psu to gpu has 1 end 8pins the other end is split into 2x 8pins, i was using one before and now two of them. i will continue to update. ill have plenty of testing to do today.
 

BugFreak

Joined
Apr 29, 2010
Location
Central FL
So you had both plugs in, just using the splitter at the end like Corsair does? I've used my 970 like that for years now without problem so I wouldn't think it is that. You made it sound like you only had one plug on the GPU plugged in and the other empty. That is not the case right?

After everything you have done so far, I definitely agree it is either drive or PSU though considering it seems to happen on your newer heavy draw games like COD. Because it is easier to swap I would start with the PSU then if that doesn't fix it a drive is needed.
 

wingman99

Member
Joined
Dec 10, 2003
thanks for inputs ! well yeah the single cable from psu to gpu has 1 end 8pins the other end is split into 2x 8pins, i was using one before and now two of them. i will continue to update. ill have plenty of testing to do today.

You were using one cable with the two 8 pin connectors when you had the problem?
 

EarthDog

Gulper Nozzle Co-Owner
Joined
Dec 15, 2008
Location
Buckeyes!
That's interesting... keep us posted for sure.

I use that config on the testing machine...