recurrent freezing of new Wintel build

Forums › Desktop computing › recurrent freezing of new Wintel build

Sideface

9349 posts

Uber Geek

Trusted

Lifetime subscriber

#318434 17-Jan-2025 10:25

Background: I am an experienced builder of small form factor Wintel desktop PCs (12 over the last 2 years), without any problems ... until now.

The new build:

Built on a test bench with a known-good PSU & GPU ... using a brand-new mini-ITX motherboard, CPU, RAM, and SSD.

Major problems with random boot failures during and after Windows 11 install - the screen and PC freezes and the CPU fan runs flat out.

This freezing occurs at any stage of boot, or shortly after a successful boot.

No warning LEDs show on the mobo.

new mobo: MSI MPG B760I EDGE WIFI DDR4 Mini ITX Gaming Motherboard (with the latest BIOS update, Sep 2024) "Supports Intel® Core™ 14th/ 13th/ 12th Gen Processors"

new CPU: Intel Core i5-14600KF Processor (gen 14)
new RAM: Corsair Vengeance LPX 32GB DDR4 (properly seated)
new SSD: Samsung 980 1TB M.2 NVMe

These one-at-a-time substitutions have all failed:

swap RAM - no fix
swap CPU (a new gen 14 Intel Core i5 14400 CPU) - no fix
swap to larger heatsink - no fix
swap SSD - no fix

Is the motherboard the problem?

All advice gratefully received. 🙄

Sideface

1 | 2

Networkn
32349 posts

Uber Geek

ID Verified

Trusted

Lifetime subscriber

#3332633 17-Jan-2025 10:38

Is the heatsink attached properly. I'd remove and reapply thermal paste. Not too much. Videos online about it explain the right amount.

nzkc

1571 posts

Uber Geek

#3332636 17-Jan-2025 10:50

Have you tried updating the firmware of the motherboard (if it has an update of course)?

lxsw20

3552 posts

Uber Geek

Subscriber

#3332639 17-Jan-2025 11:09

What happens if you boot to a bootable USB - like Ubuntu?

Andib

1363 posts

Uber Geek

ID Verified

Trusted

#3332688 17-Jan-2025 11:25

If the CPU is overheating, It will be thermal throttling which will cause Windows to freeze / crash.
In the bios under Advanced > Overclocking > CPU configuration lower the CPU lite mode. A lot of posts on reddit that MSI have this set too high by default.

<#
.DISCLAIMER
Anything I post is my own and not the views of my past/present/future employer.
#>

SpartanVXL

1306 posts

Uber Geek

#3332744 17-Jan-2025 12:20

I don’t see any overheating symptoms unless you have a temperature reading somewhere. If the machine crashes then the fans will default to 100%, this doesn’t mean it’s overheating.

Because it’s intel 14th gen you need to rule out the voltage issue that gen has. What Andib mentioned is related, check what the motherboard is pushing, disable boost or even set a very conservative manual clock and voltage so you can at least boot without issue. Then work your way back.

Sideface

9349 posts

Uber Geek

Trusted

Lifetime subscriber

#3332746 17-Jan-2025 12:53

Thank you all for your advice.

networkn Is the heatsink attached properly. I'd remove and reapply thermal paste. ... Yes, already done - I also tried a larger heatsink - no fix

nzkc Have you tried updating the firmware of the motherboard? Yes, already done, to the latest BIOS (Sep 2024) - no fix

Andib ... In the BIOS under Advanced > Overclocking > CPU configuration lower the CPU lite mode. A lot of posts on reddit that MSI have this set too high by default.

"CPU Lite Load: [Auto] > [Mode 10]" on a scale of 1 to 23, but no fix - should I go lower ... say, 5?

Sideface

mentalinc

3225 posts

Uber Geek

Trusted

#3332747 17-Jan-2025 12:58

Given all the tests does sound like a faulty motherboard.

Have you set the BIOS to the safe defaults (or what ever MSI calls it) after updating the bios?

CPU: AMD 5900x | RAM: GSKILL Trident Z Neo RGB F4-3600C16D-32GTZNC-32-GB | MB: Asus X570-E | GFX: EVGA FTW3 Ultra RTX 3080Ti| Monitor: LG 27GL850-B 2560x1440

Quic: https://account.quic.nz/refer/473833 R473833EQKIBX

Equinox IT

Cloud spending continues to surge globally, but most organisations haven’t made the changes necessary to maximise the value and cost-efficiency benefits of their cloud investments. Download the whitepaper From Overspend to Advantage now.

Andib

1363 posts

Uber Geek

ID Verified

Trusted

#3332748 17-Jan-2025 13:00

Sideface:

"CPU Lite Load: [Auto] > [Mode 10]" on a scale of 1 to 23 but no fix - should I go lower ... say, 5?

It is a performance setting so I wouldn't recommend going any lower than you have to. Start at 9 and work your way down (or drop to 5 and work your way back)

<#
.DISCLAIMER
Anything I post is my own and not the views of my past/present/future employer.
#>

djtOtago

1149 posts

Uber Geek

#3332769 17-Jan-2025 14:41

Does sound like a Motherboard issue, but I would still try installing a Linux distribution just to rule out some low level Windows hardware driver issue with that Motherboard.

Sideface

9349 posts

Uber Geek

Trusted

Lifetime subscriber

#3332799 17-Jan-2025 16:40

djtOtago:

Does sound like a Motherboard issue, but I would still try installing a Linux distribution just to rule out some low level Windows hardware driver issue with that Motherboard.

In desperation, I have just attempted to install Ubuntu (latest) via USB - another total fail - the PC stalled within seconds of the Ubuntu logo appearing.

I have now wasted 2 working days on this disaster. This motherboard hates me!

Just as well that I'm a retired geek ... if I was still a CIO this would be embarrassing! 🙄

Sideface

mentalinc

3225 posts

Uber Geek

Trusted

#3332808 17-Jan-2025 17:07

RMA time

CPU: AMD 5900x | RAM: GSKILL Trident Z Neo RGB F4-3600C16D-32GTZNC-32-GB | MB: Asus X570-E | GFX: EVGA FTW3 Ultra RTX 3080Ti| Monitor: LG 27GL850-B 2560x1440

Quic: https://account.quic.nz/refer/473833 R473833EQKIBX

SpartanVXL

1306 posts

Uber Geek

#3332809 17-Jan-2025 17:12

Yep, if it can’t run some thing like 4.5GHz at around 1.3v then sounds like a faulty motherboard given two cpus tested.

telans

13 posts

Geek

#3333112 19-Jan-2025 07:25

Sounds like the motherboard, but out of curiosity was the replacement NVMe also a Samsung or WD drive? I had a similar issue a while back I was tearing my hair out over where the computer would simply freeze soon after booting on both Linux and Windows. It turned out to be an issue with the SSD's Autonomous Power State Transition (APST). The drive reports incorrect timings on transitions between power states, or the motherboard has problems with it.

See https://bugzilla.kernel.org/show_bug.cgi?id=208123

I was able to fix/work around it in Linux with the kernel parameter nvme_core.default_ps_max_latency_us=0. Not sure how to disable it on Windows. Unlikely this would be your issue but it's something to test.

Azzura

603 posts

Ultimate Geek

ID Verified

#3333117 19-Jan-2025 08:13

what about the Extreme Memory Profile (XMP) settings?

Sideface

9349 posts

Uber Geek

Trusted

Lifetime subscriber

#3333246 19-Jan-2025 14:51

Thank you all for your suggestions.

The story so far:

I have again swapped out the M.2 SSD, this time for a known-good Corsair SSD with Windows 11 pre-installed (borrowed from another Wintel PC).

I now have an intermittently functional Windows 11 OS [NOT yet activated]. It still freezes at random intervals during or after reboot. The screen freezes and the mouse arrow disappears.

The longest time between crashes is 5-10 minutes (even at idle), so the new PC remains useless.

I have established that the PC is not overheating - typical CPU temperatures are in the 40-50 Centigrade range.

To make things more complicated, I am now getting two error messages:

SecurityHealthSystray: SecurityHealthSystray.exe - Bad Image
\\?\C\Windows\System32\SecurityHealth\10.0.27703.1006-0\SecurityHealthSSO.dll is either not designed to run on Windows or it contains an error.
Error status 0xc000012f

and

A driver cannot load on this device.
Driver: cpu149_x64.sys
A security setting this as a vulnerable driver and blocking it from loading. You'll need to adjust your settings to load this driver. *

* because the PC is not yet activated, I can't get into Core isolation page. 😒

Sideface

1 | 2