Question Games Crash/Freeze - 3 year old PC, recent issue

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Specs:
CPU: i9900ks
GPU: MSI 2080ti
RAM: 16GB trident z
Board: Hero Max XI
PSU: Seasonic prime ultra 750 w
Storage: C:\ = Evo 970 pro (1TB), D:\ = Corsair MP510 (2tb)
OS: Windows 10 21H2


What's happening?

Starting playing Dragon Age: Inquisition - all was well until 40-60 hours in the game started freezing and needed to be 'end processed'. This was happening once every 6-8 hours at first, but after a couple weeks of continuing to play it was happening every 3-5 minutes. Sometimes it would pop up a directX error.

Underclocking GPU made it work - until my PC shutdown (Yeah UNDERclocking crashed my PC).

I thought it was a DAI specific issue until I opened up Battletech, which I just played 100 hours of with no issues and it also froze and crashed. COD: Warzone also freezes and crashes (this one to a directX error). Heroes of the storm plays fine, CS:GO plays fine.


What I've done:
Uninstalled/Reinstalled graphics drivers
Run games in various compatibility modes
Change display settings in windows and NVidia control panel
Removed all overlays (Not sure I had any on to begin with, but I checked).
Repair attempts on steam/origins
vcredist_x64 and x86 used to repair/uninstall/reinstall
fullscreen/windowed/etc.
RAM seated properly
PSU Cables re-done (and there are 3 individual cables from the PSU to the GPU)



A bunch of other stuff from google searches and forum searches here.


Oddity - I've noticed on games that are likely to crash, in-game buttons (ACCEPT for example) will need to be clicked twice before they work.


Since my PC worked just fine for almost 3 years, I assume that a reformat would solve whatever problem, but it's a huge PITA and the problem may either return or persist.

I believe it's a directX issue, but I don't know where it would have come from. The only thing I remember having issues with at the time these crashes started was trying to play Elden Ring - which I ended up refunding because of the stuttering.

Any help would be appreciated.
 
Last edited:
Do you have access to 3dmark? running a 20 loop stress test may give some answers.

If you do not have that OCCT has a decent free stress test program that can test the vram and other things on the gpu.

May give us some idea of what is going on.

Or it might just be a power supply issue.
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
I also ran the Power test - but my CPU spiked to 100C and I cancelled. Attempted a few, all stopped at about 20 seconds.
Looks like I just ran the power test incorrectly - Re-did it using SSE, will post results when it's complete (not burning up CPU now).


CPU test temps stay around 63-69C.
 
Last edited:

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Things seem to be getting worse.

Even while playing Heroes of the Storm, which is usually cake, my computer screen will go black, everything will freeze for a few seconds and then come back.

After effects of these mini crashes pictured below

 
Reactions: George Stamper

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
This problem is growing rapidly. I'm not sure what to check now or what it might be stemming from. Anything that requires use of the GPU makes the screen flicker black and freeze the program (usually a game).

At this point, I'm not even sure the nuclear option (reformatting) would actually have an effect.
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Removed my Audio Card and used Mobo audio - problems persisted (recommended by activison)

Did a clean uninstall/reinstall of video drivers (DDU).

Warzone got worse - took longer to load, constant stuttering, crashed at basically the same point (haven't even been able to get passed tutorial). No DirectX error this time though.

Battletech - graphics will freeze after a certain point, but the game will continue running (I can switch between units).

DAI - same issues of crashing.

HOTS - some black screening, but continues running (Black screen flickering happening more often).
 
Reactions: George Stamper

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Been going down rabbit holes and tangents of research trying to figure this problem out.

My surge protector is 3 years old and has likely seen some abuse (Shared a circuit with a portable AC unit last year).

Could a bad power strip make it seem like a bad PSU? I've already swapped the PSU with no change in issues so I put the seasonic back (80 PLUS Titanium vs Gold for the new one, wanted to stick to the better unit).
 
Been going down rabbit holes and tangents of research trying to figure this problem out.

My surge protector is 3 years old and has likely seen some abuse (Shared a circuit with a portable AC unit last year).

Could a bad power strip make it seem like a bad PSU? I've already swapped the PSU with no change in issues so I put the seasonic back (80 PLUS Titanium vs Gold for the new one, wanted to stick to the better unit).
I wish I could be more of a help but it seems your issue is going to be a pain to find. Kinda like my poor wifi connection which turned out to be RF interference from my video card. Sometime just sucks and takes time to find the issue.

Could be anything but with all your trial and error maybe the CPU, but I don't think so because you ran OCCT and the program does a very good job of finding issues with CPU, Ram GPU etc.

Doesn't happen often but could be the motherboard, as everything on it's own is running correctly, maybe it just throws a error once in awhile which messes things up.

I have only ever had one bad motherboard and it would randomly crash the pci lanes for the GPU, took me forever to figure it out RMA'd the board and boom issues all went away.
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
How did you find the motherboard issue? I've never seen an error change from AO(kay).

I've ran stress tests for several hours with no issues, so I'm lost as to why games will crash, since they don't exactly stress my system. Is it worth paying for better benchmark software to test longer (OCCT only allows 1 hour at a time)?

I loaded Warzone in 'safe mode' and managed to play through it this morning.

Battletech also ran just fine today, until I let it sit for a few minutes, then it froze, but I got a solid 2.5 hours of play time.
 
How did you find the motherboard issue? I've never seen an error change from AO(kay).

I've ran stress tests for several hours with no issues, so I'm lost as to why games will crash, since they don't exactly stress my system. Is it worth paying for better benchmark software to test longer (OCCT only allows 1 hour at a time)?

I loaded Warzone in 'safe mode' and managed to play through it this morning.

Battletech also ran just fine today, until I let it sit for a few minutes, then it froze, but I got a solid 2.5 hours of play time.
I didn't find it per say. I rma'd the video cards thinking that was more likely. I was sent new ones and the issue still persisted so I then RMA'd the board. After getting a new replacement board back "same model" my issues were gone.

I will only put around 8 total hours into troubleshooting before it becomes a waste of my time and I just sell everything and start over anymore lol.
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
The fact that stress test are all passing along with windows running. Possible your SSD is going back and only comes up when loading gb up gb of textures for a game like warzone.

Also. Have you did a fresh install of Windows 10 since all of these issues started?

I consider reinstalling windows my 'nuclear' option since I don't want to take the time to backup 3 years worth of tinkering.

Atm the only thing that's kept me from doing it is my windows install USB is in a box in a storage pod somewhere and I'm a few months from moving it all back out. I'll have to buy a new copy of windows 10 :'( . (I didn't save the key).
 
I consider reinstalling windows my 'nuclear' option since I don't want to take the time to backup 3 years worth of tinkering.

Atm the only thing that's kept me from doing it is my windows install USB is in a box in a storage pod somewhere and I'm a few months from moving it all back out. I'll have to buy a new copy of windows 10 :'( . (I didn't save the key).
Generally windows 10 keys save, I can't member how it works but if you use the iso, Say I don't have a key windows update will usually reattached your old key.

https://www.microsoft.com/en-us/software-download/windows10
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Just for s&g go to userbenchmark.com and run the test. Post your results here. Make sure to disable any overlay such as Afterburner before running.
UserBenchmarks: Game 176%, Desk 100%, Work 176%
CPU: Intel Core i9-9900KF - 98%
GPU: Nvidia RTX 2080-Ti - 184.3%
SSD: Samsung 970 Evo Plus NVMe PCIe M.2 1TB - 309%
SSD: Corsair Force NVMe PCIe M.2 2TB - 224%
RAM: G.SKILL Trident Z RGB DDR4 3200 C16 2x8GB - 70.9%
MBD: Asus ROG MAXIMUS XI HERO (WI-FI)


Generally windows 10 keys save, I can't member how it works but if you use the iso, Say I don't have a key windows update will usually reattached your old key.

https://www.microsoft.com/en-us/software-download/windows10
I'll go for a reformat this weekend.

Currently everything is running without crashing. Only change I've made is a new power strip.

Hoping all my problems weren't because of a fried strip, but if that's starting to stabilize, I might as well handle the software side of it too since the games are still a bit choppy.
 
UserBenchmarks: Game 176%, Desk 100%, Work 176%
CPU: Intel Core i9-9900KF - 98%
GPU: Nvidia RTX 2080-Ti - 184.3%
SSD: Samsung 970 Evo Plus NVMe PCIe M.2 1TB - 309%
SSD: Corsair Force NVMe PCIe M.2 2TB - 224%
RAM: G.SKILL Trident Z RGB DDR4 3200 C16 2x8GB - 70.9%
MBD: Asus ROG MAXIMUS XI HERO (WI-FI)




I'll go for a reformat this weekend.

Currently everything is running without crashing. Only change I've made is a new power strip.

Hoping all my problems weren't because of a fried strip, but if that's starting to stabilize, I might as well handle the software side of it too since the games are still a bit choppy.
Make sure RAM is installed in slots a2-b2. Be sure to enable XMP in BIOS.
 

omnilicious

Distinguished
Dec 7, 2011
51
7
18,535
0
Make sure RAM is installed in slots a2-b2. Be sure to enable XMP in BIOS.
XMP should have been enabled already - but it looks like my bios went back to Auto for some reason.

UserBenchmarks: Game 187%, Desk 105%, Work 190%
CPU: Intel Core i9-9900KF - 101.3%
GPU: Nvidia RTX 2080-Ti - 186.5%
SSD: Samsung 970 Evo Plus NVMe PCIe M.2 1TB - 324.8%
SSD: Corsair Force NVMe PCIe M.2 2TB - 213.2%
RAM: G.SKILL Trident Z RGB DDR4 3200 C16 2x8GB - 98.2%
MBD: Asus ROG MAXIMUS XI HERO (WI-FI)




Tentatively calling this solved. New powerstrip seems to have made the biggest difference - went with a more robust one Tripp-Lite Iso.

Enabling XMP since it likely switched off when I reseated the ram before also helped smooth out choppiness I was noticing.

A freaking power strip issue :'( .
 
Last edited:

ASK THE COMMUNITY

TRENDING THREADS