Question Games Have Begun Crashing, Trying to Diagnose What Part is Dying

[Aliased]

Games have begun running much lower FPS than normal, trying to diagnose what's going wrong.

I've already done a clean install of the gpu drivers using seperate software that fully wiped them first, forgot the name, as well as a rollback to older drivers.

HWiNFO logs here, I've highlighted when I *think* the crash occurred as well as hidden a lot of other rows, let me know if I should unhide: https://docs.google.com/spreadsheets/d/1LfFTK55Xa4wFy_QE_iSblpHm0we8a78ZvVUM_uTtUxA/edit?usp=sharing

I also have event logs, is there a way to do an attachment?

OS: Win11 version 10.0.22631 Build 22631
CPU: AMD Ryzen 7 5800X3D
MOBO: TUF Gaming B550M-Plus (WI-FI)
GPU: nvidia GTX 3080 Founders Edition
Ram: Crucial Ballistix RGB 3600 MHz DDR4 DRAM 32GB (16gb x 2)
PSU: RM750 - Can't recall for certain if it's new to this build, 4 years. Or from my last build, probably more like 8 years.

Drives:
Corsair Force Series MP600 1TB Gen4 PCIe X4 NVMe: 25GB free (Windows on this drive)
Samsung 970 EVO Plus SSD 2TB: 1.5TB free
Samsung 850 EVO 250GB: 75GB free
ST3000DM008-2DM166 3TB: 230GB free (HwIiNFO is telling me this has unrepairable sections, so I might need to replace that soon. No programs are installed on it, strictly video storage.)

Screenshots from two seperate days: today, where I had a few crashes and gave up on gaming, and last week when I tried a bunch of things to solve the issue and kept crashing over and over:

Had to get a friend to upload the images, idk why.

 

[Ralston18]

Update your post to include full system hardware specs and OS information.

= = = =

Include PSU: make, model, wattage, age, condition (original to build, new, used, refurbished).

Disk drive(s): make, model capacity, how full?

Instead of Event Viewer (for now anyway) look in Reliability History/Monitor.

Use the Day view.

Expand the window so all can be seen and select a date that seems to have the most errors of any sort but focus on red circles.

Take a screenshot of the window and post the screenshot here via imgur (www.imgur.com).

 

[Aliased]

Take a screenshot of the window and post the screenshot here via imgur (www.imgur.com).

Still working on the rest, but imgur isn't letting me upload. I grab a link and then try to follow the link and it says 404. If I try to post publically it says SHARE_POST_FAIL.

Edit: fixed by the power of friendship.

 

[Ralston18]

On the imgur screen do you see the green "+ New Post" icon in the upper left corner?

Click the green icon and drop the image(s) in the dash walled box.

Then when finished, copy the resulting URL and post that URL here.

 

[Aliased]

On the imgur screen do you see the green "+ New Post" icon in the upper left corner?

Click the green icon and drop the image(s) in the dash walled box.

Then when finished, copy the resulting URL and post that URL here.

yeah, I do that and then the post 404's when I try to use the link. And trying to post to community to see if that worked resulted in that error.

Like I said, fixed with the power of friendship. I sent the images to a buddy and he was able to post. I can't get anything to upload. It shows as uploaded in my account but I can't go look at it myself even.

Regardless, images are now up and linked.

 

[Aliased]

On the imgur screen do you see the green "+ New Post" icon in the upper left corner?

Click the green icon and drop the image(s) in the dash walled box.

Then when finished, copy the resulting URL and post that URL here.

Have I otherwise added everything needful to the post? Any specs I didn't know to include?

 

[Ralston18]

Screenshots very helpful.

Increasing numbers of errors and varying errors is a sign of a faltering/failing PSU.

And that is what I am seeing in the screenshots. View the technical details for specific codes. The codes may or may not be helpful.

Referring back to:

"PSU: RM750 - Can't recall for certain if it's new to this build, 4 years. Or from my last build, probably more like 8 years."

The PSU may be nearing or at its' designed in EOL (End of Life).

If possible swap in another known working, well rated PSU with equal or higher wattage. Be sure to use only the cables that come with that PSU.

Be sure that you have backed up all important data and proven that the data is recoverable and readable.

Likely there are corrupted files that will need to be dealt with.

 

[Aliased]

Screenshots very helpful.

Increasing numbers of errors and varying errors is a sign of a faltering/failing PSU.

And that is what I am seeing in the screenshots. View the technical details for specific codes. The codes may or may not be helpful.

Referring back to:

"PSU: RM750 - Can't recall for certain if it's new to this build, 4 years. Or from my last build, probably more like 8 years."

The PSU may be nearing or at its' designed in EOL (End of Life).

If possible swap in another known working, well rated PSU with equal or higher wattage. Be sure to use only the cables that come with that PSU.

Be sure that you have backed up all important data and proven that the data is recoverable and readable.

Likely there are corrupted files that will need to be dealt with.

The day with a lot of errors was me fussing with driver installation/uninstallation and settings and safe mode and shutting the computer off via the power button and such. I think a lot of those errors are things like "windows didnt' start (because I held the power button rather than let it finish booting because I missed the safe mode prompt.) I shared that day because you said the day with the most errors, but I'm fairly sure most of those were caused by my own messing around trying to diagnose/fix the problem in a number of different ways a friend advised me to try.

See below for examples of the actual errors that are happening consistently. They're all exclusively nvlddmkm.sys errors.

https://imgur.com/a/pBeScRY

View: https://imgur.com/a/pBeScRY

 

[Aliased]

Be sure that you have backed up all important data and proven that the data is recoverable and readable.

Likely there are corrupted files that will need to be dealt with.

Regarding this, are you saying *if* it is a faltering PSU that a faltering PSU corrupts files? Or was that comment seperate from the PSU issue, just a blanket statement that corrupted files happen?

 

[Ralston18]

Power losses be they internal (PSU) or external (black out or brown out) can and do cause file corruption.

A faltering PSU causing power losses within the host computer can prevent the OS from properly finishing up its' "housekeeping" necessary for the next boot up. Just as if power was cut to the entire PSU.

Unfortunately, a festering PSU problem may not be immediately noted and all ends badly.

That is why so much emphasis is placed on backing up data.

And also not working inside the computer while power is on - that can be even more damaging with sparks, smoke, and fire.

Turn off and unplug before doing anything.

 

[Aliased]

Makes sense, will do a backup before I do anything.

Did you check the new screenshots? Do you still think PSU when all the errors say nvlddmkm?

Does it change your opinion at all if I further clarify that when I say games are "crashing" what is actually happening is the screen is freezing but I can still hear combat or menu options going on if I move the controls? It's *only* the graphics freezing, until I have to force close the game and then it shows up as a game crash on the event viewer.

 

[Aliased]

Power losses be they internal (PSU) or external (black out or brown out) can and do cause file corruption.

A faltering PSU causing power losses within the host computer can prevent the OS from properly finishing up its' "housekeeping" necessary for the next boot up. Just as if power was cut to the entire PSU.

Unfortunately, a festering PSU problem may not be immediately noted and all ends badly.

That is why so much emphasis is placed on backing up data.

And also not working inside the computer while power is on - that can be even more damaging with sparks, smoke, and fire.

Turn off and unplug before doing anything.

Not sure how forums work, if I needed to reply for you to see my post. See above questions.

 

[Ralston18]

Reference: "Do you still think PSU when all the errors say nvlddmkm?"

Yes.

The PSU is a critical component.

In order to work properly the PSU must receive steady input power that is within the required specifications for voltage and current.

Likewise the PSU must output voltages (3.3, 5, and 12) to various system components. Those output voltages must also be within the required voltages and currents needed by the served components. And knowledgeable end users can do some testing to check voltages.

FYI:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

So a problem with one voltage can create problems while devices or components using the other voltages are unaffected. So graphics fails and audio remains.

The end results of any given power glitch can and do vary. If, for example the computer is reading or writing to a disk drive, a power glitch interrupts the process and the file either loses data or the data is changed (corrupted). Errors result. The system may or may not be able to recover from the errors.

And, if not able to recover, then you see:

"A problem with your hardware caused Windows to stop working correctly".

Which is a good indication that Windows files, drivers files, and application files are being corrupted in some manner that disrupts system performance or from even operating at all (BSOD).

Windows may or may not be able to accurately report on what happened.

Sometimes, if the problem is a specific file that has been corrupted, the file can be repaired or reinstalled.

And all works again.

Until the next glitch.

Increasing numbers of varying errors are an indication of a faltering or failing PSU.

However, if the same file keeps getting corrupted then the cause is something else.

 

[Aliased]

Reference: "Do you still think PSU when all the errors say nvlddmkm?"

Yes.

The PSU is a critical component.

In order to work properly the PSU must receive steady input power that is within the required specifications for voltage and current.

Likewise the PSU must output voltages (3.3, 5, and 12) to various system components. Those output voltages must also be within the required voltages and currents needed by the served components. And knowledgeable end users can do some testing to check voltages.

FYI:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

So a problem with one voltage can create problems while devices or components using the other voltages are unaffected. So graphics fails and audio remains.

The end results of any given power glitch can and do vary. If, for example the computer is reading or writing to a disk drive, a power glitch interrupts the process and the file either loses data or the data is changed (corrupted). Errors result. The system may or may not be able to recover from the errors.

And, if not able to recover, then you see:

"A problem with your hardware caused Windows to stop working correctly".

Which is a good indication that Windows files, drivers files, and application files are being corrupted in some manner that disrupts system performance or from even operating at all (BSOD).

Windows may or may not be able to accurately report on what happened.

Sometimes, if the problem is a specific file that has been corrupted, the file can be repaired or reinstalled.

And all works again.

Until the next glitch.

Increasing numbers of varying errors are an indication of a faltering or failing PSU.

However, if the same file keeps getting corrupted then the cause is something else.

Would HWiNFO pin down for certain if it it's the PSU? I know it has sensors for the voltage, I don't know if it's possible for the PSUs potential issue to be undetectable to those sensors?

 

[beyondlogic]

Games have begun running much lower FPS than normal, trying to diagnose what's going wrong.

I've already done a clean install of the gpu drivers using seperate software that fully wiped them first, forgot the name, as well as a rollback to older drivers.

HWiNFO logs here, I've highlighted when I *think* the crash occurred as well as hidden a lot of other rows, let me know if I should unhide: https://docs.google.com/spreadsheets/d/1LfFTK55Xa4wFy_QE_iSblpHm0we8a78ZvVUM_uTtUxA/edit?usp=sharing

I also have event logs, is there a way to do an attachment?

OS: Win11 version 10.0.22631 Build 22631
CPU: AMD Ryzen 7 5800X3D
MOBO: TUF Gaming B550M-Plus (WI-FI)
GPU: nvidia GTX 3080 Founders Edition
Ram: Crucial Ballistix RGB 3600 MHz DDR4 DRAM 32GB (16gb x 2)
PSU: RM750 - Can't recall for certain if it's new to this build, 4 years. Or from my last build, probably more like 8 years.

Drives:
Corsair Force Series MP600 1TB Gen4 PCIe X4 NVMe: 25GB free (Windows on this drive)
Samsung 970 EVO Plus SSD 2TB: 1.5TB free
Samsung 850 EVO 250GB: 75GB free
ST3000DM008-2DM166 3TB: 230GB free (HwIiNFO is telling me this has unrepairable sections, so I might need to replace that soon. No programs are installed on it, strictly video storage.)

Screenshots from two seperate days: today, where I had a few crashes and gave up on gaming, and last week when I tried a bunch of things to solve the issue and kept crashing over and over:

Had to get a friend to upload the images, idk why.

i would replace the psu

have you changed anything major wise.
have you updated from windows 10 to 11 etc ( recently)