[SOLVED] PC crashes to black screen for every game ?

[Zinni]

It just started yesterday and it has progressively gotten worse today.

It's a brand new fresh build and I've been playing a bunch of games with no problems for a couple of weeks now.

Yesterday I installed the Resident Evil 2 remake from Steam. I was playing it for maybe 15 to 30 minutes and all of a sudden the screen goes black. I thought the monitor shut off because I could still hear everything in the game. I tried to move around but I didn't hear my footsteps. There was nothing I could do to get the screen to come back on so I figured it crashed and I did a hard shutdown.

I booted the PC up again and did a little research and a lot of people seem to have had crashing issues with the game. I implemented various recommended fixes, such as shutting Ray Tracing off, and started the game again, played for 15 or so minutes and it crashed. This time the fans kicked in full blast. I hard shutdown the PC, booted it back up and started the game again and was able to play for over an hour with no problems.

Today, I started playing the game again and it crashed in 5 mins with fans at full speed. Started playing again and it crashed just as quickly, but this time the fans did not go full speed. Tried again and it crashed again. I figured this game is a dud.

I tried another Steam game, Conan Exiles. Played for 20-30 minutes and THAT GAME crashed to black. I tried Friday the 13th and that game crashed within 10 mins. I thought that maybe it's a Steam issue, so I booted up an Epic Games Launcher game and that game crashed to black as well.

Before playing Resident Evil, yesterday, I had played Friday the 13th for well over an hour with no problems. Is it possible that Resident Evil corrupted something? I reinstalled Steam. I reinstalled my GPU drivers. I must note that the day BEFORE yesterday there were two power outages, while the PC was running, due to a storm. But I was able to play Friday The 13th after those outages, but before Resident Evil, with no problems.

I checked Event Viewer and the only critical error I got was from hard shutting the computer down. I checked Reliability Monitor and it says I had a hardware error for each crash. I tried to do a minidump and it would never create a minidump folder.

I generated a log from GPU-Z and for only one of the crashes it said the GPU and memory temps reached 70 degrees at the time of the crash. For the rest of the crashes they barely even hit 60.

My BIOS is up to date along with all my drivers. Windows is activated and updated.

PC build:
MSI PRO Z790-A WIFI DDR5 ATX Motherboard
Intel Core i9-13900K (No OC)
MSI RTX 4090 Gaming Trio (No OC)
CORSAIR Vengeance 32GB (2 x 16GB) RAM (XMP on) (Window diagnostic RAM check, no problems)
(x2) SAMSUNG 970 EVO PLUS M.2 2280 500GB (OS) 2TB (Storage) (health checked: 100%)
Fractal Design Meshify 2 Mid Tower Computer Case
NZXT - Kraken X73 RGB All-in-one 360mm Radiator CPU Liquid Cooling System
MSI MPG A1000G PCIE 5.0 1000W PSU
LG Ultragear 27" 240hz 1440p OLED gaming monitor
Windows 10 Home

Any recommendations on how to resolve this would be greatly appreciated.

Thank you!

 

[Ralston18]

My thought is that something has come loose inside the computer due to expansion/contraction and/or vibrations.

Where did you get the computer? Does it have any warranties?

If you are comfortable doing so (or know someone who is) power down, unplug, open the case.

Verify that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place.

Look for signs of damage: bare conductor showing, melted insulation, pinched or kinked wires, blackening or browning, swollen components, corrosion,
loose or missing screws. Something stuck in the fan blades.

Anything that seems amiss....

 

[Aeacus]

Any recommendations on how to resolve this would be greatly appreciated.

If the Ralston18 fixes doesn't help, then there are 2 other steps, neither of the two is cheap:

1. Try with 2nd, known to work GPU.
2. Try with far beefier PSU.

#1 It could be that power outages damaged your GPU, hence the black screens. While unlikely that GPU would be an issue, it still is a possibility.

#2 For RTX 4090, minimum PSU wattage would be 1200W, due to GPU transient power spikes. While 1600W PSU is preferred. E.g any of these 1.6kW "kings",
pcpp: https://pcpartpicker.com/products/compare/cJbwrH,2BcRsY,tFvdnQ/

Bonus:
Get an UPS. With that expensive hardware, UPS is a must. Since UPS will keep your PC running when blackouts appear. This avoids OS corruption or the very worst, hardware failure due to power loss.

Good UPS brands to go for are CyberPower, APC and Tripplite. Topology would be line-interactive and output waveform would be true/pure sine wave. <- Can explain further as of why that topology and output waveform.

 

[Zinni]

My thought is that something has come loose inside the computer due to expansion/contraction and/or vibrations.

Where did you get the computer? Does it have any warranties?

If you are comfortable doing so (or know someone who is) power down, unplug, open the case.

Verify that all connectors, cards, RAM, jumpers, and case connections are fully and firmly in place.

My friend, you are the genius of yourself.

I built the PC with my slightly more detailed PC building friend. We built it at his house and I drove it home. And then we had an earthquake last week. And then the temperature started fluctuating. It would be warm one day so I'd turn the heat off and wake up to it being in the high 30s with the room freezing.

I opened the computer up this morning. I checked all the cable connections and they were nice and snug. The GPU was also snug, though MSI provides a pretty poorly designed brace to hold this behemoth up. It goes along the side of the card and they give you these little pieces of rubber to stick between the brace and the card. Well those pieces of rubber had fallen into the PSU cables at the bottom, which is not the easiest to access with this case design. I got a proper brace coming tomorrow and I jammed some cardboard in there for the time being.

I then started pushing on the RAM, the thing that Windows said was all good, and lo and behold there was a 'click'. Surely this couldn't be it.

I put it all back together and loaded up a game. I'm 5 mins in. 10 mins in. 25 MINS IN?!? I was able to complete the match. So then I decided to test it with the very scary (and not for its content this time) Resident Evil 2...with ray tracing on.

That was 3 hours ago. The only time the screen went black was when I paused the game and was on the phone for too long and the monitor went to sleep (set by the monitor, not by Windows).

Thank you! You're intuition is excellent.

I'm gonna keep this unsolved for right now, just in case something else might still be wrong, but I think I'm in the clear.

 

[Ralston18]

Hopefully so.

Not uncommon. No one wants to force anything especially with a new build.

You were careful about what you did and paid attention thereafter. That is what matters.

No harm, no foul.

Fingers crossed.

 

[zx128k]

Was going to say black screen when game starts implies power cable but someone look to have got there first.

 

[Zinni]

#1 It could be that power outages damaged your GPU, hence the black screens. While unlikely that GPU would be an issue, it still is a possibility.

#2 For RTX 4090, minimum PSU wattage would be 1200W, due to GPU transient power spikes. While 1600W PSU is preferred. E.g any of these 1.6kW "kings",
pcpp: https://pcpartpicker.com/products/compare/cJbwrH,2BcRsY,tFvdnQ/

Bonus:
Get an UPS. With that expensive hardware, UPS is a must. Since UPS will keep your PC running when blackouts appear. This avoids OS corruption or the very worst, hardware failure due to power loss.

Good UPS brands to go for are CyberPower, APC and Tripplite. Topology would be line-interactive and output waveform would be true/pure sine wave. <- Can explain further as of why that topology and output waveform.

I've had mixed feelings towards UPSs. I love the concept but from having a relatively extended period of time (for one job) of having to rely on a group of APCs, I got turned off to the idea. They were very finicky and somewhat unreliable (it was for a recording studio on a bus). But after doing some research (inspired by your reply) I realize now that they were old and overworked.

Where I live it barely rains, but for these last few months we've been getting tons of it. In the six years I've lived in this neighborhood I've experienced 1 power outage. Now I've had two in one day. I ordered two CyberPower UPSs (one for my work setup and one for my gaming/entertainment setup). They put out a pure 60hz sine wave when switched over to the battery, where the APCs put out a mangled waveform. It's supposedly not that big of a deal, especially since I'm only using them to protect my stuff as opposed to continuing to work off the battery power, but there's barely a difference in price between the two companies.

As for the PSU, I'm gaming at 1440p and my card is in line with the founder's edition in terms of power consumption, and transient spikes were only 30 - 40% higher according to Gamers Nexus' testing (though he did say that something might make it spike higher, they just didn't see anything more in their 8 hours of testing). I'm also not OCing the GPU or CPU. Is it likely that I could get spikes above 1000 watts?

 

[Aeacus]

Where I live it barely rains, but for these last few months we've been getting tons of it. In the six years I've lived in this neighborhood I've experienced 1 power outage. Now I've had two in one day. I ordered two CyberPower UPSs (one for my work setup and one for my gaming/entertainment setup). They put out a pure 60hz sine wave when switched over to the battery, where the APCs put out a mangled waveform. It's supposedly not that big of a deal, especially since I'm only using them to protect my stuff as opposed to continuing to work off the battery power, but there's barely a difference in price between the two companies.

UPS isn't only for those locations with constant power issues. UPS is actually for everywhere, and the more expensive hardware you have - the more reasons to get an UPS.

For example;
In the place where i live, we, very rarely, get any issues from main electricity grid. Still, nothing isn't 100% solid and when the last blackout happened (lasted only few seconds, but well enough to shut down our PCs at an instant), i lost 2 hours worth of work. At that point, it was enough for me and i went and bought our UPSes. Since then (i've used UPSes for ~5 years now), UPS has kicked in 1-2 times per year.

While our UPSes have kicked in very rarely, but even 1 sudden power off (if we wouldn't have UPS) can cause all sorts of issues. Starting from loss of any unsaved work, continuing on to data corruption, or the very worst, killing our PSU or whole PC. <- This is something i won't take my chances on, since you can not predict electricity grid issues, to safely shut down your PC beforehand. Predicting where lightning strikes has better chances, than predicting main electricity grid issues.

APC UPSes, for the most part, yes, output simulated sine wave. This isn't good for sensitive electronics, which PCs are. Hence why use true/pure sine wave output for PCs.

As for the PSU, I'm gaming at 1440p and my card is in line with the founder's edition in terms of power consumption, and transient spikes were only 30 - 40% higher according to Gamers Nexus' testing (though he did say that something might make it spike higher, they just didn't see anything more in their 8 hours of testing). I'm also not OCing the GPU or CPU. Is it likely that I could get spikes above 1000 watts?

It's true that Steve (GamersNexus) didn't saw as high transient power spikes with RTX 40-series, as they were with RTX 30-series. But Steve also kept the chance open, that it might spike higher.

Here, question then remains: Would you take your chances or would you completely void any chances?

With that kind of expensive hardware, i, personally, would remove any chances. And go with 1.6 kW PSU instead.
Getting an UPS is essentially with the same idea, to avoid any chances of your software/hardware issues. Since without UPS, there is a chance that blackout can kill your hardware. That, and good UPS + PSU combined, still cost less than new RTX 4090 (if the blackout happens to kill the GPU).

 

[Zinni]

Hopefully so.

Not uncommon. No one wants to force anything especially with a new build.

You were careful about what you did and paid attention thereafter. That is what matters.

No harm, no foul.

Fingers crossed.

Aaahh...well unfortunately it was not the fix. In fact it black screened/crashed while watching Youtube today. It was so bad that Nvidia completely disappeared off my computer (aside from GeForce Exp). Nvidia Control Panel was gone. I could only output out of the motherboard's HDMI via integrated graphics.

GeForce Exp. told me that a new driver was available, even though it was the same one I installed yesterday. I proceeded to install it again but it failed. I forgot where I looked next, but whatever it was told me there was no Nvidia GPU in my system, even though it was plugged in with all its RGB lights going (no fans though).

I pulled out the 4090, grabbed my previous 3080ti, popped out the CMOS battery and put it back in for good measure, and installed the 3080ti. I would've tested the 4090 in my old rig but the case is too small. The 3080ti worked. I was able to install the drivers. I pulled it out and put the 4090 back in and that worked, and I proceeded to install the drivers for that. Played a game for a while.....black screen, so I reseated both RAM sticks. Then I pulled up a Youtube video about Memtest86 and it black screened again.

I ran Memtest86 for the full 13x4 test and it passed with no errors.

So right now I'm thinking (and hoping) it's CRU. LG is weird and for some reason some games (Unreal Engine games and supposedly other engines too) become 4k in fullscreen. So my 1440p monitor displays a warning that the resolution is 4k when it should be 1440, regardless how windows or the in game settings are set and it also cuts you down from 240hz to 120hz. Doesn't happen with Call of Duty which is the id Tech engine and doesn't happen with Red Dead 2 which is RAGE. And it doesn't happen in Windows Borderless for any game either. CRU was the only thing that fixed this.

However the CRU developer has stated that people have been getting black screens when using CRU with newer Nvidia drivers, and he doesn't know why. The suggested fix is to DDU all GPU drivers, clean install the Nvidia drivers and then run CRU again.
I just DDU'd Nvidia and Intel and only installed Nvidia. I didn't reapply CRU. I want to test some games without it first, even with the LG 4k upscale/downscale mess, just to see if it black screens.

I'll update with the results.
I'll also post the weird errors I'm seeing in Event Viewer. I'm getting repeated networking errors (even though my internet is seemingly running fine) along with the expected display related errors.

 

[Ralston18]

Start with Reliability History versus Event Viewer.

Reliablity History is much more end user friendly and the timeline format can be revealing.

Also download all drivers manually directly via the applicable manufacturer's website. Reinstall and reconfigure as necessary. No third party tools or installers. You keep control - not them.

Pay close attention for any check boxes etc. that may default to additional installations and configurations that you probably would not want.
Read carefully and be sure to fully scroll all presented windows even if the desired option(s) are being directly presented.

 

[Zinni]

It just happened again and it wiped the Nvidia drivers as well. I was in a game but alt-tabbed and watching Twitch at the same time. I did not reapply CRU, so my game opened in the center of my screen this time, about 1/4 sized, with the 4k incorrect resolution message from the monitor. I switched to Windowed Fullscreen, it filled my screen and I just kept it like that. Before it crashed I was playing for at least and hour.

Earlier in the day the Microsoft Store decided to reinstall the Intel graphics drivers, even though I had set it to "No" in the Device Installation Settings. I don't know where else to stop these auto-updates because I don't have the gpedit.msc capability.

Anyway, I used the integrated graphics output to boot into Safe Mode. Once in safe mode it defaulted to using my GPU as the signal source. I first deleted all monitor profiles from CRU and deleted the app as well. I DDU'd Nvidia and Intel's drivers and did another clean install of the latest Nvidia drivers, and the card is working again in regular Windows.

Besides the entire PC itself, here are some additional new things to my setup, and most of them were introduced about the same time as those initial power outages.

• I have a new ISP. I went from AT&T slow DSL (they lied to my face and said it was fiber, but at least it was full duplex) to Spectrum 1gb cable.
• I have a new wireless mouse. I went from a Razer Viper Ultimate (also wireless) to a Razer V2 Pro. The new mouse comes with it's own wireless USB transmitter. I also uninstalled all of Razer's software a few days ago.
• I have been using a Yamaha AG03 audio interface for years. The latest driver comes with an app that doesn't work right. You apply settings and they don't stay. The old version of the app that I used on my previous PC isn't compatible with the new drivers, so I have older drivers installed at the moment.
• Intel GPU drivers. I never saw Intel GPU drivers when looking in my previous PC's Device Manager.

Error messages:
Reliability Monitor doesn't really show me much.

Common Reliability Monitor messages:
Critical Events:

• Windows - Hardware error.
• GameManagerService3 - Stopped working
• Windows was not shut down properly
• Desktop Window Manager - Stopped working
• dwm.exe - Stopped working
• NVIDIA Control Panel Application, 8.1.940.0 - Stopped working (only one instance)
• NVIDIA Container - Stopped working (only one instance)
• MSI.CentralServer - Stopped working (only one instance)

Event Viewer messages: From Windows Logs > Systems
Warnings: (these are from today before and after the crash)
- Source: DistrbutedCOM. Event ID: 10016 (this one shows up a lot with variations to the descriptions between 7:29pm and 7:34pm)

Spoiler: Full description

The machine-default permission settings do not grant Local Activation permission for the COM Server application with CLSID
{C2F03A33-21F5-47FA-B4BB-156362A2F239}
and APPID
{316CDED5-E4AE-4B15-9113-7055D84DCC97}
to the user DESKTOP-(then it shows my name and some numbers) from address LocalHost (Using LRPC) running in the application container Microsoft.Windows.ShellExperienceHost_10.0.19041.1949_neutral_neutral_cw5n1h2txyewy SID (more numbers. not sure if they're personal or not). This security permission can be modified using the Component Services administrative tool.

- Source: DistrbutedCOM. Event ID: 10016 (here's a specific one that repeats a lot)

Spoiler: Full description

The application-specific permission settings do not grant Local Activation permission for the COM Server application with CLSID
{6B3B8D23-FA8D-40B9-8DBD-B950333E2C52}
and APPID
{4839DDB7-58C2-48F5-8283-E1D1807D0D7D}
to the user NT AUTHORITY\LOCAL SERVICE SID (S-1-5-19) from address LocalHost (Using LRPC) running in the application container Unavailable SID (Unavailable). This security permission can be modified using the Component Services administrative tool.

- Source: e2fexpress. Event ID: 27 (7:31pm)

Spoiler: Full description

The description for Event ID 27 from source e2fexpress cannot be found. Either the component that raises this event is not installed on your local computer or the installation is corrupted. You can install or repair the component on the local computer.

If the event originated on another computer, the display information had to be saved with the event.

The following information was included with the event:

Intel(R) Ethernet Controller (3) I225-V

The message resource is present but the message was not found in the message table

- Source: WLAN-AutoConfig. Event ID: 10002 (7:31pm)

Spoiler: Full description

WLAN Extensibility Module has stopped.

Module Path: C:\Windows\System32\DriverStore\FileRepository\netwtw6e.inf_amd64_ddded575f3a87757\IntelIHVRouter12.dll

- Source: Kernel-PnP. Event ID: 219 (7:25pm)

Spoiler: Full description

The driver \Driver\RtkUsbAD_2330 failed to load for the device USB\VID_0DB0&PID_D1D7&MI_00\6&1a5ac4f7&0&0000.

- Source: Kernel-PnP. Event ID: 219 (7:25pm)

Spoiler: Full description

The driver \Driver\ysusb_w10_64 failed to load for the device USB\VID_0499&PID_170D&MI_00\7&72544be&0&0000.

- Source: e2fexpress. Event ID: 27 (7:25pm)
Same description as before

- Source: Kernel-PnP. Event ID: 219 (7:25pm) (3 of them at the same second)

Spoiler: Full descriptions

The driver \Driver\NVHDA failed to load for the device HDAUDIO\FUNC_01&VEN_10DE&DEV_00A3&SUBSYS_14625103&REV_1001\5&192422e8&0&0001.

The driver \Driver\nvvad_WaveExtensible failed to load for the device ROOT\UNNAMED_DEVICE\0000.

The driver \Driver\Netwtw12 failed to load for the device PCI\VEN_8086&DEV_2725&SUBSYS_00248086&REV_1A\4&2bb179bf&0&00E2.

- Source: WLAN-AutoConfig. Event ID: 10002 (7:24pm)

Spoiler: Full description

WLAN Extensibility Module has stopped.

Module Path: C:\Windows\System32\DriverStore\FileRepository\netwtw6e.inf_amd64_ddded575f3a87757\IntelIHVRouter12.dll

- Source: DistrbutedCOM. Event ID: 10016 (7:07pm)

Spoiler: Full description

The application-specific permission settings do not grant Local Launch permission for the COM Server application with CLSID
Windows.SecurityCenter.WscDataProtection
and APPID
Unavailable
to the user NT AUTHORITY\SYSTEM SID (S-1-5-18) from address LocalHost (Using LRPC) running in the application container Unavailable SID (Unavailable). This security permission can be modified using the Component Services administrative tool.

There's a bunch more DistributedCOMs and a e2fexpress between 7:04 and 7:07.

Critical:
- Source: Kernel-Power. Event ID: 41 (7:04pm)
The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly.

Errors:
- Source: DistrbutedCOM. Event ID: 10005 (there were 217 of these from 7:25pm to 7:29pm)
There's multiple instances of these descriptions.

Spoiler: Full Description

DCOM got error "1084" attempting to start the service RmSvc with arguments "Unavailable" in order to run the server:
{581333F6-28DB-41BE-BC7A-FF201F12F3F6}

DCOM got error "1084" attempting to start the service ShellHWDetection with arguments "Unavailable" in order to run the server:
{DD522ACC-F821-461A-A407-50B198B896DC}

DCOM got error "1084" attempting to start the service TokenBroker with arguments "Unavailable" in order to run the server:
Windows.Internal.Security.Authentication.Web.TokenBrokerInternal

DCOM got error "1084" attempting to start the service GoogleChromeElevationService with arguments "Unavailable" in order to run the server:
{708860E0-F641-4611-8895-7D867DD3675B}

DCOM got error "1084" attempting to start the service BthAvctpSvc with arguments "Unavailable" in order to run the server:
{6E1F7F3E-760E-45F3-AA8F-5761ABDA272A}

DCOM got error "1084" attempting to start the service camsvc with arguments "Unavailable" in order to run the server:
Windows.Internal.CapabilityAccess.CapabilityAccess

DCOM got error "1084" attempting to start the service BITS with arguments "Unavailable" in order to run the server:
{4991D34B-80A1-4291-83B6-3328366B9097}

DCOM got error "1084" attempting to start the service WSearch with arguments "Unavailable" in order to run the server:
{E48EDA45-43C6-48E0-9323-A7B2067D9CD5}

DCOM got error "1084" attempting to start the service WpnUserService_47da3 with arguments "Unavailable" in order to run the server:
{1FFE4FFD-25B1-40B1-A1EA-EF633353BB4E}

DCOM got error "1084" attempting to start the service EventSystem with arguments "Unavailable" in order to run the server:
{1BE1F766-5536-11D1-B726-00C04FB926AF}

DCOM got error "1084" attempting to start the service camsvc with arguments "Unavailable" in order to run the server:
Windows.Internal.CapabilityAccess.CapabilityAccess

I probably missed a couple. A vast majority were the TokenBroker ones.

- Source: EventLog. Event ID: 6008
The previous system shutdown at 7:02:19 PM on ‎3/‎3/‎2023 was unexpected.

These are pretty much what my events look like regularly. All the Criticals, Errors and Warnings are repeated with a vast majority being DistributedCOMs.

Here's a couple that have shown up a bunch before but not today.
- Source: Service Control Manager. Event ID: 7023
The NVIDIA LocalSystem Container service terminated with the following error:
A generic command executable returned a result that indicates failure.
- Source: Service Control Manager. Event ID: 7031
The NVIDIA LocalSystem Container service terminated unexpectedly. It has done this 1 time(s). The following corrective action will be taken in 6000 milliseconds: Restart the service.

Administrative Events shows a bunch of different Warnings (Event IDs not already included in the list above: 122, 360, 63, 200, 202, 201, 6000, 1) but most of the Errors are DistributedCOMs and unexpected system shutdowns.

I guess my next move is to install the 3080ti and see how that handles. Or maybe I should do a fresh install of Windows first. I am tempted to take this to a repair shop, though I don't know who to trust.

 

[Zinni]

For anyone that might stumble upon this thread, I solved the problem.

It was the GPU's power cable. I swapped it out with a replacement and it hasn't black-screened in weeks.
I can't believe all the hoops I jumped through and it ended up being the stupid cable.
In my profession, the cables are the first things you check when something is going wrong. Why didn't I apply that to this?