[SOLVED] Random freezing and I can't find out if the issue is software- or hardware-related

[Im_Ben]

Specs:

• i7 - 8700k
• Asus Strix 2080 OC
• Asus Prime Z370-A
• Samsung 970 EVO Plus (windows installed here)
• Samsung 860 SSD
• Corsair Dominator RGB DDR4 3200MHz 32GB
• Corsair RM650x 650W (v2)

I'm struggling with random freezes, and I can't locate if the issue is hardware- or software-related. I've re-installed windows 10, re-flashed the BIOS, re-installed GPU driver (with DDU first). The freezes are random, everything from watching a youtube video to gaming. However, when trying to re-flash the BIOS it kept freezing over and over again, and it did not recover. When the pc usually freezes it recovers, but not when it's in the BIOS.

I ran memtest86, and the RAM passed, however, when I went into my BIOS the pc froze immediately and did not recover. Currently running AIDA64 to stress the CPU.

I noticed these issues about a week ago but did not install or change anything to the pc-config or anything software related.

I have not OC the GPU nor the CPU. I have not re-installed MSI afterburner after I did a fresh install of Windows.

 

[octavecode]

You only use Driver Verifier to check your drivers.If you system boots fine then you disable it again.
Always check for stability with system running stock,do not run any OC
Also keep in mind that XMP profiles don't always work as intended , it depends on the setup.

Test your system stock with fresh windows, after a week or so enable XMP and after a couple of weeks check your OC.

 

[Ralston18]

How old is that 650 Watt Corsair PSU? Heavy use for gaming, graphics work, or even bit-mining?

Look in Reliability History and Event Viewer for error codes, warnings, and even informational entries that correspond with the freezes.

Right-clicking any entries will provide more information. Likely to be cryptic and/or confusing.

File History may be worth a look as well.

Start with Reliability History - user friendly. The timeline format can be very helpful. I.e., what things happened a week ago when the issues started.

Event Viewer takes more time and effort to both navigate and understand. However, you may well find a clue or two therein.

Also: try reinstalling the GPU drivers manually - no DDU.

Go directly tot the GPU manufacturer's website to download. Manually reinstall and configure the GPU drivers.

 

[Im_Ben]

How old is that 650 Watt Corsair PSU? Heavy use for gaming, graphics work, or even bit-mining?

Look in Reliability History and Event Viewer for error codes, warnings, and even informational entries that correspond with the freezes.

Right-clicking any entries will provide more information. Likely to be cryptic and/or confusing.

File History may be worth a look as well.

Start with Reliability History - user friendly. The timeline format can be very helpful. I.e., what things happened a week ago when the issues started.

Event Viewer takes more time and effort to both navigate and understand. However, you may well find a clue or two therein.

Also: try reinstalling the GPU drivers manually - no DDU.

Go directly tot the GPU manufacturer's website to download. Manually reinstall and configure the GPU drivers.

I actually almost gave up when I got no answers here but was so happy when I received the notification that someone had replied, thank you! After I posted my questions, I've tried to replicate the issue, but with no success. However, noticed that when running AIDA 64 that Windows freezes when trying to put 'more' stress with moving around the AIDA 64 window like crazy, and it kinda did freeze, not in the normal 4-6 seconds freezes, but the 1-second freeze. I do not know if that is related to the issue I'm having, or just the process being delayed since the CPU is being stressed at 100% utilization.

1. The PSU is from Black Friday 2018, you could say for heavy gaming and some small graphics work.
2. Two things stood out to me:
   1. The first one: Error setting traits on Provider {8444a4fb-d8d3-4f38-84f8-89960a1ef12f}. Error: 0xC0000001 (with some reports matching the freezing, but not everyone) (16 errors in the last week, but this is a fresh install of windows)
   2. The second one:
   3. The application-specific permission settings do not grant Local Activation permission for the COM Server application with CLSID
      {2593F8B9-4EAF-457C-B68A-50F6B8EA6B54}
      and APPID
      {15C20B67-12E7-4BB6-92BB-7AFF07997402}
      to the user XXXXXX-XXXXXXX\XXXXXX SID (S-1-5-21-1179249972-1252989738-2014924459-1001) 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 (205 warnings in the last week, but is matching some of the timestamps of the freezing, I blanked out the user just to be on the safe side )
3. No luck with file history
4. No luck with manually installing the GPU drivers, and then installing them again from the manufacturer.

EDIT: The window freezing when 'shaking' it only happens with AIDA64.

 

[Ralston18]

Try running the built in Windows 10 Troubleshooters. The troubleshooters may find and fix something.

Next try "sfc /scannow" and "dism".

References:

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161

https://www.howtogeek.com/222532/how-to-repair-corrupted-windows-system-files-with-the-sfc-and-dism-commands/

The idea being to find and fix any file corruption. Get the errors down and then continue to watch to determine if they come back, how many, how fast, and what they are.

 

[Im_Ben]

Try running the built in Windows 10 Troubleshooters. The troubleshooters may find and fix something.

Next try "sfc /scannow" and "dism".

References:

https://www.lifewire.com/how-to-use-sfc-scannow-to-repair-windows-system-files-2626161

https://www.howtogeek.com/222532/how-to-repair-corrupted-windows-system-files-with-the-sfc-and-dism-commands/

The idea being to find and fix any file corruption. Get the errors down and then continue to watch to determine if they come back, how many, how fast, and what they are.

I did those two tests, but got the message saying everything was good/passed.

 

[Ralston18]

Use either Task Manager or Resource Monitor to observe system performance. Use both but only one at a time.

Open the window and slide to one side or the other.

Observe system behavior while the system idles, while you browse or do light work. Then while gaming and lastly while stress testing via AIDA64 etc..

What resources are being used, to what extent (%), and what is using the resource.

Watch for what changes and happens keeping in mind that you are expecting a freeze.

It may take a few tries to discover some pattern or set of events. Be methodical and try to vary only one thing at a time.

 

[Im_Ben]

Use either Task Manager or Resource Monitor to observe system performance. Use both but only one at a time.

Open the window and slide to one side or the other.

Observe system behavior while the system idles, while you browse or do light work. Then while gaming and lastly while stress testing via AIDA64 etc..

What resources are being used, to what extent (%), and what is using the resource.

Watch for what changes and happens keeping in mind that you are expecting a freeze.

It may take a few tries to discover some pattern or set of events. Be methodical and try to vary only one thing at a time.

I got a small stutter right now, and when looking at resource monitor, the CPU dropped all the way to zero. It might be the issue here, what are your thoughts on that? Or is it the PSU/mb that isn't delivering the power correctly?

http://imgur.com/a/5qmIqwK

View: https://imgur.com/a/5qmIqwK

 

[Ralston18]

Readling back:

A 2+ year old PSU should be okay but there are no guarantees especially with respect to how much demand what actually put upon it over those years.

And how "good" that PSU was to begin with can be questioned: design, components, assembly, etc... Hopefully it was new - that can sometimes be in question.

PSU is a suspect. Yet, there are still some things you can do:

1.) Continue the monitoring. Watch for something else that happens at the same time.

2.) Power down, unplug, open the case. Clean out dust and debris. Reseat all cards, connectors, RAM, and jumpers. Ensure that all are firmly in place.

3.) Go back to Reliability History and Event Viewer. Look for an error entry that may correspond with the above noted stutter or other noted Monitor events.

For the most part the numbering is a bit arbitrary so you can actually do the three things in the order that is easiest or more convenient to you. Or prioritize as you deem appropriate.

Suggested reading:

https://www.tomshardware.com/reviews/best-psus,4229.html

Use two or three of the PSU wattage calculators to assess your system's power requirements. Do a manual wattage total as well. If a component has a wattage range then use the high end value.

How much difference between the calculated wattage and the 650 watts provided by the Corsair RM650x? (My personal preference is a an extra 25% margin.)

 

[Im_Ben]

Readling back:

A 2+ year old PSU should be okay but there are no guarantees especially with respect to how much demand what actually put upon it over those years.

And how "good" that PSU was to begin with can be questioned: design, components, assembly, etc... Hopefully it was new - that can sometimes be in question.

PSU is a suspect. Yet, there are still some things you can do:

1.) Continue the monitoring. Watch for something else that happens at the same time.

2.) Power down, unplug, open the case. Clean out dust and debris. Reseat all cards, connectors, RAM, and jumpers. Ensure that all are firmly in place.

3.) Go back to Reliability History and Event Viewer. Look for an error entry that may correspond with the above noted stutter or other noted Monitor events.

For the most part the numbering is a bit arbitrary so you can actually do the three things in the order that is easiest or more convenient to you. Or prioritize as you deem appropriate.

Suggested reading:

https://www.tomshardware.com/reviews/best-psus,4229.html

Use two or three of the PSU wattage calculators to assess your system's power requirements. Do a manual wattage total as well. If a component has a wattage range then use the high end value.

How much difference between the calculated wattage and the 650 watts provided by the Corsair RM650x? (My personal preference is a an extra 25% margin.)

I've been doing quite a few tests lately, all from PSU to CPU.

• 90% of when the pc freezes it's something with the CPU utilization.
  • It either drops to 0% or spikes to 100%(that only happened ones).
  • When monitoring the resource monitor, I've noticed that the "System Interrupts" process jumped to the top.
    • Upon researching it, It seems like it either it some driver issue (so I disabled every device in device manager to check if it went away, and it did not, also disabled something in the network adapter). From what other people have been saying about that process; is that it should not be a problem unless it's over 20+ percent. Mine is running at 0.1%, however, when turning all the other cores off beside one, it's at 17%. So... a bit hard to get the full picture if that process is really the cause of the issue or not.
  • I cleaned every single bit of the case, re-fitted all the cables, and looked for external damages on components, and it looked basically brand new (There's usually not much dust, I tend to do regular maintenance on it, to clear out the dust etc.). Also cleared the CMOS, just to be sure.
  • Nothing new has come up in the event viewer or reliability monitor.
  • I did as you asked went to different PSU wattage calculators and they more or less said that 500W should be more than enough.
    • I also read up on the reviews of the PSU, and they were good (I bought the PSU brand new, but if it was refurbished unit or something else, the store did not provide that information, but I'm like 99% sure that it's brand new).
    • I also got myself a "wattage monitor" (a very interesting piece of tech, quite fun to monitor it), and under 100% full load on the CPU and GPU it hit 390W. That's without XMP(I turned it off the just see if it was the XMP that's causing the issue) or overclocking on the CPU or the GPU. I can't really imagine that with OC, the PSU will pull 250W more.

After all this, it still seems to be happening.

 

[Ralston18]

Another tool:

Process Explorer. Download from Microsoft.

https://docs.microsoft.com/en-us/sysinternals/downloads/process-explorer

Try to observe what else is going on when the "System Interrupts" jumps to the top or the CPU goes to 0% as noted in the graph (Post #7).

 

[Im_Ben]

Another tool:

Process Explorer. Download from Microsoft.

https://docs.microsoft.com/en-us/sysinternals/downloads/process-explorer

Try to observe what else is going on when the "System Interrupts" jumps to the top or the CPU goes to 0% as noted in the graph (Post #7).

Nothing really stands out when the freezing happens, beside the CPU behaviour. I'm going, to be honest, the freezing is starting to become extremely annoying. I'm willing to buy a new PSU and a new CPU to see if that fixes it, and I can always return them. What are your theories here? What component are you leaning towards being the issue here?

 

[octavecode]

You can try Windows Driver Verifier and see if there is faulty driver but from what i'm reading the board looks off.

 

[Im_Ben]

You can try Windows Driver Verifier and see if there is faulty driver but from what i'm reading the board looks off.

I ran it but looks like it didn't find anything wrong. Can I ask you why you think the board is the issue here?

 

[octavecode]

Random crashes usually mean there is a power delivery issue.Some times its the PSU and sometimes the VRM of the board is faulty and you already tested CPU and RAM.
Your PSU in decent and not that old so.....
IF you can get a PSU from a friend for a couple of days and do some testing then do it but i bet my money on the board.You also said you have some bios problems?

 

[Im_Ben]

Random crashes usually mean there is a power delivery issue.Some times its the PSU and sometimes the VRM of the board is faulty and you already tested CPU and RAM.
Your PSU in decent and not that old so.....
IF you can get a PSU from a friend for a couple of days and do some testing then do it but i bet my money on the board.You also said you have some bios problems?

Yes, I've had the BIOS freeze on me a few times when I tried to re-flash it, however, it has not happened again since.

 

[octavecode]

Have you tried downgrading the bios? Which version are u running?

 

[Im_Ben]

Have you tried downgrading the bios? Which version are u running?

I'm running the latest bios, however, that's from 2019. They stopped updating that motherboard, I'm guessing since they come out with a rev 2 version.

I have not tried to downgrade the bios.

 

[octavecode]

Try to downgrade to the first bios ver that came with the board and test your system for any random crashes and also disable spectre/meltdown fixes with inspectre

 

[Im_Ben]

Try to downgrade to the first bios ver that came with the board and test your system for any random crashes and also disable spectre/meltdown fixes with inspectre

Another tool:

Process Explorer. Download from Microsoft.

https://docs.microsoft.com/en-us/sysinternals/downloads/process-explorer

Try to observe what else is going on when the "System Interrupts" jumps to the top or the CPU goes to 0% as noted in the graph (Post #7).

I have not been ignoring the replies, however, the issue has not come back (in 5 days).

So I ran Windows Driver Verifier again, and the whole system felt off, It was really slow, and long and behold in task manager "system interrupts " was running at 80%. My theory was there had to be some driver issues, but I already re-installed windows 10 as a fresh install (as mentioned in a previous post) but got the issues when I ran Windows Driver Verifier again. So, therefore, I tried to install a fresh copy of windows from a non-windows machine, because when running windows' media creation to create a win10 USB, and might just maybe copy over something from the current windows 10 build that's already installed (I might be completely wrong here), but nonetheless, it has not happened since. I really hope that has fixed it, if not I'll send an update. I'm also checking if maybe some of the OC was also the issue here. I ran the system 2-3 days without any OC, and have now turned on XMP, and afterwards, I'm turning on the OC for the CPU etc.

 

[octavecode]

You only use Driver Verifier to check your drivers.If you system boots fine then you disable it again.
Always check for stability with system running stock,do not run any OC
Also keep in mind that XMP profiles don't always work as intended , it depends on the setup.

Test your system stock with fresh windows, after a week or so enable XMP and after a couple of weeks check your OC.