Question Very strange freezing issue ?

[tuffghost]

Ok, I'll try to keep this as short as possible, but here's what's been going on. A few weeks ago I went to restart my system and Windows froze before exiting, forcing me to hard reset. I didn't think much of it, but Windows was buggy and unresponsive after the reboot, so I restarted again. Once again it froze before going through with the restart, but then from there it didn't boot. Most of the time it didn't get past BIOS at all, occasionally I could get as far as the Windows login screen but then it would fully hang. Not having much in the way to test hardware, especially without a computer to make boot disks from and the like, I got it diagnosed by a PC repair shop. They tested CPU, PSU, GPU, SSDs, etc. on a testbench and narrowed it down to the motherboard. And so, I bought a new motherboard, rebuilt the system, and...

After about a day of full stability, the issue is back.

With this motherboard I do have a little more stability, I can occasionally get the thing running, but here's the extremely weird thing to me: the common factor before every single freeze is that it's when the system is relatively idle. If I, for example, have a YouTube video playing, or have a benchmark or test running, or am playing a game, the system will run. But once I stop? It will freeze within maybe 30 seconds. This seems to me like it might indicate that there's something weird going on with the voltages when the CPU starts idling, or something to that effect, but I'm not knowledgeable enough in this sort of thing to be able to detect it. The voltages seem normal in HWiNFO64, but since I can't see what's happening the moment of the freeze who can say for sure. I will say that one of the CCD sensors is always a good 10–15℃ hotter than the other one, but I don't know, maybe that's just the 5950X for you.

Most of the time the freezing occurs without a crash so I'm not even given a dumpfile. Or it happens before getting to Windows at all. Every now and then I've gotten a BSOD, but nothing that gives a whole lot of information. The failing driver is always ntoskrnl.exe, which as I understand it doesn't shed much light on the true problem. There's nothing of use on the mobo debug code.

Here's what I have done:
I've updated the BIOS, I've reverted the BIOS to older versions. I've installed and uninstalled and reinstalled chipset drivers. I have run memtest86+ and testmem5 for multiple passes without issue (testmem5 occupying the CPU with something to do meant I got to leave the system running for hours). I've uninstalled and reinstalled any drivers I could think of and unplugged my audio interface, DAC, PCI capture card, every other USB device I have. I've turned PBO fully off in the BIOS, I've kept memory settings stock instead of clocking them to XMP speeds.

So, what do I do at this point? It seems highly unlikely to me that two motherboards would be bad in the same exact way, so I'm mostly ruling that out. The fact that the freezing occurs in many cases before ever booting into Windows, or even when I tried booting into a Linux live USB, rules out a boot drive issue or some sort of drivers issue. This leads me to think that the problem is with the CPU or PSU, though as I understand it PSUs typically don't fail in that kind of way. The computer has otherwise been just fine to me for the past two years.

And finally, the specs:

• AMD Ryzen 9 5950X
• ASUS ROG Crosshair VIII Dark Hero (orig. mobo was a Gigabyte AORUS X570 Master)
• G.Skill 2x16GB F4-3600C16D-32GTZRC | Trident Z RGB DDR4-3600 CL16-19-19-39 1.35V
• Gigabyte Vision 3080
• 2 WD SN850 NVMe SSDs
• 1 8TB HDD I don't remember
• Seasonic FOCUS GX-850, 850W 80+ Gold PSU

Thank you so much to anyone who helps out. I'm losing my mind here, this has all been so frustrating and disheartening and I don't have the money to just start wildly dumping into possible solutions right now. Getting halfway to just becoming an "off the grid" kind of freak by this point.

 

[Darkbreeze]

Try disconnecting ALL of the drives, by literally removing them completely from the M.2 slots and the SATA cabling. Disconnect power from the SATA drive as well.

See if you are now able to consistently POST and access the BIOS.

If not, and the problem remains, then my next suggestion would be to beg, borrow or steal (Unless you already have one) ANY other graphics card that is known working, and swap it out for your 3080. This should tell us if it's a drive or the graphics card, or if we need to look at it being a different issue. I'd discount ANYTHING the shop told you, since clearly they were not correct about it being the motherboard, so all their results are suspect and in fact if you paid them to diagnose the issue I'd go back and ask for a refund since the motherboard clearly was not the problem.

And if they kept your other motherboard I'd go ask for it back too because there is probably nothing wrong with it. Also, if you have ANY other USB devices connected to the PC other than the mouse and keyboard, I'd disconnect those as well until you resolve the issue. Faulty USB devices can commonly cause a variety of seemingly unrelated problems that ARE in fact related, but just don't seem like they would be.

As far as the PSU not being the problem because PSU problems "don't fail in that way", umm, no, totally not true. ANY and ALL problems that you could ever see or experience on any computer can ALWAYS be the result of the power supply because a weak or faulty power supply can mimic any other problem since ALL hardware relies on the power supply and if there is a problem with the power supply it can SEEM as though that component is faulty when in fact it's actually the PSU making it misbehave. Keep in mind there are several different power rails on a power supply, including potentially multiple 12v rails, 5v rail, 3v rail, etc., and different devices may use different parts of the PSU so while mostly the PSU might act like it's fine, a problem with any one part of the PSU might make one of your components flake out or simply not work.

 

[stonecarver]

You wrote a lot so if I missed it, it seems you have not reinstalled windows. I'm going back thinking when you said all was good with computer and on a shut down forced manual is where it went down hill.

Could have been installing a major update and with out knowing it you corrupted windows.

Though you said you also tried a Linux live USB. So ?

You also said your just becoming an "off the grid" kind of freak by this point. Does that mean your bugging out or using a inverter. Just asking if it plays out knowing how to help fix PC issue if your on an inverter.

I would check windows file system.

Open command prompt and run as administrator

Type "sfc /scannow" into the command prompt that opens and press enter. Let finish


Dism /Online /Cleanup-Image /CheckHealth

Dism /Online /Cleanup-Image /ScanHealth

Dism /Online /Cleanup-Image /RestoreHealth

let each one finish before you start next. How is your system before and after if it finds and fixes files.

restart when your all done than test

 

[tuffghost]

Haha, no that part was a joke, as in, the frustration of all this is making me want to give up technology.

In any case, I have now since tried all the following:

• A new PSU
• Disconnecting every combination of drives and USB devices possible. Freezing was occurring even without the Windows boot drive installed so that at least rules out it being a Windows problem.
• Using both DIMMs individually
• I remembered I still had a 1080 lying around so even though I was fairly certain it wasn't a GPU problem I ensured it wasn't.

Which all strongly hints that it is a CPU problem after all. I don't know what else it could possibly be other than the replacement motherboard being broken in the same exact way as the previous one (unlikely) or I suppose, theoretically, some kind of disastrous goings on with the electricity in the house.

Going to try to RMA the 5950X, though AMD really makes you jump through a lot of hoops for the process, and then I'll report back. Unless anyone's still got any ideas floating around.

 

[Darkbreeze]

CPU is the next step, although, WHAT is the EXACT model of the "new PSU" you purchased?

Very unlikely, in fact, nearly impossible to have the exact same problem with a new motherboard IF the motherboard was the problem. So yeah, almost certainly it's not the board. Which is all the more reason to hang onto your original board or get it back if possible, if the shop kept it. And, all the more reason to seek a refund since they didn't actually do anything, since the board was not the problem.

I'd be interested, if I were you, on knowing EXACTLY how did they "test" the CPU and other hardware. Did you watch them test it. Did you actually SEE them remove the CPU, install it in another motherboard and run stress tests on it?