#Faulty Mobo, CPU, or both?

1 messages · Page 1 of 1 (latest)

turbid bone
#

Hey guys sorry for re-opening this but the troubleshooting is ongoing.

Let me explain what happened from the beginning.
So this pc was working fine until the owner supposedly plugged in an old external SSD with windows installed on it. She said it was crashing ever since (BSOD). She went ahead and replaced her NVMe thinking it was the issue but nothing changed. The PC was still under warranty so she took it to the shop and they told her she had faulty RAM so they replaced her RAM kit and told her it was fixed.

When she took it back home the BSODs came right back.

After numerous clean and fresh windows installation attempts she called me and asked me to troubleshoot it for her.

I started by performing a BIOS update followed by yet another clean and fresh windows installation to make sure everything was as clean as possible. I then updated windows / drivers and all of the typical stuff and even refrained from bloating the system with ASRock auto driver and Polychrome RGB crap.

At first it was running fine until it BSOD again. At that point I tweaked the BIOS and disabled CPU C-State Support and driver auto update. To my surprise it was fixed which then made me go to the next step replacing her CPU (i7-13700F) with a known working one I have for testing (i3-12100F). The PC was running fine with both manual and default BIOS settings for a while until it crashed again. I then noticed that even though I updated the BIOS and had installed a clean fresh windows, her static pink RGB theme persisted. I removed the CMOS battery, left it out for 5 mins and back in, loaded default BIOS settings and only enabled XMP. It was now magically fixed and had passed cinebench and furmark stress tests and was also left on idle for an hour + and on light duty (youtube videos) for another hour or so with no crashes.

Curiously enough I downloaded ASRock Polychrome again and as soon as I changed from RGB puke to static pink boom, BSOD…
I then uninstalled that crap but I got a BSOD again with both CPUs.

So I’m now wondering whether the motherboard has possible firmware-level corruption or an unstable onboard RGB controller, and whether or not her CPU is also semi-fried or not.

For reference, the BSOD codes are:
Mostly - CRITICAL_PROCESS_DIED
Rarely - KMODE_EXCEPTION_NOT_HANDLED

timber raven
#

What I would suggest, is to redo the bios, pop the battery walk away pet the cat or whatever, then reinstall windows to get rid if the bug problem from the rgb software, then test again make sure it's stable, then use a off motherboard rgb controller to remove the offending software, and try again, it's sounds like whatever the software is accessing is causing the problem. Corsair makes rgb controllers, but the software might do the same thing, if it does it might be time to swap the mobo out, and see if the mobo has issues.

turbid bone
#

And it mostly crashes on idle or light use 🥲

timber raven
#

yeah it is odd indeed, luckily the rgb on mine is easily disabled in the bios

#

go back to a known good configuration

turbid bone
timber raven
#

use a little bios change as possible aka only use xmp and reinstall windows fresh and DO not install any motherboard brand software * aka crate or other*

#

only install drivers as needed and make sure it's stable, then set a restore point, then try various software and see which one causes the crashes most likely it's the rgb software because of low level access

turbid bone
#

I even disabled fast startup etc

deep vault
#

"stress tests" of an hour or so each IN THEORY should find most problems, but not necessarily. Get it back to the stage BEFORE you "downloaded ASRock Polychrome again" (you said it was not crashing then) and run more like 24 hours of stress test. For example, run Furmark at a low res (like 640x480) while also running Furmarks "CPU Burner" using perhaps half the cores of the CPU, and just leave it like that for a day or so.

timber raven
#

well it is a 13700... they are on the list of bad cpu's from degradation, it's very liekly the bios was never updated the cpu is basically junk

turbid bone
timber raven
#

the tell tale is your 12xxx doesn't have the problem

#

that is enough to tell me the 13700 is bad

turbid bone
deep vault
#

tend to agree. If she'd been running this 13700K and had not done the BIOS updates last year when the degredation issue really came to light, it's probable her CPU got toasted.

#

Once it's damaged, there's NO going back.

turbid bone
timber raven
#

you could try an rma but good luck, you could replace it, but who knows what the new one will do and ebaying a 13/14th gen is just russian roulette waiting to happen. better to find her a 12700k and accept the performance loss

turbid bone
timber raven
#

same difference

deep vault
turbid bone
deep vault
#

It's a shitty situation Intel put everyone with 13th/14th in, and as was called out back when this all flared up the impacts would not be seen by most until long after the issue was "out of the headlines".

#

Though TBF, Intel themselves had absolutely no way to contact directly everyone that potentially affected CPU.

turbid bone
#

its very unfortunate

deep vault
#

She should dig up her receipts and TRY to go through Intel directly for an RMA.

turbid bone
#

For some reason my gut still tells me its both cpu and mobo

#

because the moment i installed the rgb software from asrock my known working cpu also crashed

deep vault
#

it's always POSSIBLE both are trouble.

#

is this a pre-built, or she bought parts individually?

turbid bone
#

how can I confirm it 100% tho?

turbid bone
#

so part by part

deep vault
#

well, ASSUMING it's a legitimate seller, her receipts for the parts SHOULD at least get her a new CPU from Intel. Couldn't say wrt Asrock. That's IF the shop itself won't replace both.

turbid bone
#

its a legit shop here in my area but i think her 2 year warranty is about to run out xD

deep vault
#

🤷‍♂️ well, when it gets to that point it's totally a YMMV. All she can do is TRY.

turbid bone
#

true

#

Oh well, Ive tried at least

deep vault
#

yep.
As for the Asrock board, I mean, if it's crashing even with your 12100F (a VERY unstressful CPU) that otherwise works fine for you, yeah I'd be trying to get a replacement mobo too. No idea how long is Asrock warranty, I don't buy their stuff much. Think I've bought only 1 Asrock product in at least the past 10 or 15 years.

turbid bone
#

yeah so with my 12100F its stable until I download ASRock software at which point it crashes

#

and then it crashes again until I reset BIOS and force stop the rgb

#

very weird

deep vault
#

okay, but will the board run your 12100F for like 24 hours straight under a real stress?

turbid bone
#

I could try that

deep vault
#

Thing is, even a maxxed load 12100F doesn't put anywhere NEAR the stresses on a mobo as a 13700. totally different animals really.

turbid bone
#

it ran for like 4-5 hours

turbid bone
deep vault
#

you could try the 24 hours stress I mentioned above with Furmark and CPU Burner (but let it use all cores, not half).

turbid bone
#

So furmark + cinebench?

deep vault
#

Furmark's "CPU Burner"

#

just click the button in Furmark and it opens it.

#

it'll use as many cores as you set, from 1 to max.

turbid bone
#

oh okay!

deep vault
turbid bone
#

thanks!

deep vault
#

it's pretty intense, so be careful with it.
it pushes with "120W TDP" 9900X to 225W.

image.png
turbid bone
#

I dag a bit and calculated 2 years timeframe and looks like she indeed had a motherboard with a bugged bios version (probably if my calculations are correct xd)

#

so yeah I'm gonna probably go ahead and deem the cpu degraded for sure

deep vault
#

They were ALL bugged.
Intel "claimed" it was "fixed" back in Sept/Oct 2024 (microcode update 0x12B), any BIOS before then was potentially (very likely) degrading CPU's.
And then they went and released ANOTHER "fix" as recently as May 2025. (microcode update 0x12F)

#

From Tom's article:
"Intel fixed the majority of Raptor Lake's instability issues in September 2024 with mega firmware update 0x12B, which addressed several major issues causing chaos on affected chips (and lead many chips to an accelerated death). Update 0x12B combined microcode updates 0x125 and 0x129, and addressed elevated voltage requests by Raptor Lake CPUs during light activity or when idle."

#

Since you said she'd never updated the BIOS since originally buying it, yeah, betcha her CPU is dead.
"Prior to these firmware updates, many Raptor Lake chips were experiencing accelerated wear and tear, especially on the K-series parts — due to extreme stress on the clock tree circuit inside the IA core, and excessively high voltages and temperatures. Failure of the clock tree circuit would cause a shift in the clock duty cycle, leading to system instability — and, in many cases, CPU death."

turbid bone