Random rebooting

domquark

Hi All,
About 15 months ago, we bought two SEi12-1235U Beelink PC’s, both with 32GB of RAM and 1TB NVMe SSD’s. All was good at first, but both have developed the same fault.
Both PC’s now reboot randomly.
I have tried:
• Reset the BIOS (front button) – no effect
• Reset the BIOS to factory default – no effect
• Changing the “State after G3” BIOS setting – neither S0 or S5 setting has any effect (the PC always reboots)
• Changing the power supply
• Changing the RAM (including different modules with different speeds)
• Using only one RAM module in different slots
• Cleaning and replacing the heatsink compound on the CPU cooler
• Changing the NVMe SSD for brand new drives
• No errors are being logged in the Event Viewer (apart from unexpected restart errors)
• The reboot is completely random, but always occurs within 30 minutes of turning on – sometimes reboot will occur before Windows has fully loaded. One PC started rebooting every few hours, but the reboots have steadily become more frequent.
• Running a heavy load program has no effect – reboot still occurs randomly – so the issue appears not to be load/heat related
Both PC’s will run quite happily continuously (for hours) without rebooting when left in the BIOS.
Both PC’s are running the latest BIOS (109).
Neither shows any signs of CPU overheating (both are below 50C when reboot occurs).
One curious thing is that the original SSD’s seem to last longer before the system reboots, compared to the new drives.
Are they dead or is there something else that I can try? Any help greatly appreciated.

Beelink CS-George

1.Test with Linux

Boot from a Ubuntu Live USB. If stable for >1 hour, focus on Windows/drivers; if unstable, suspect hardware.

2.Minimal Boot in Windows

Use msconfig to disable all non-Microsoft services and startup items. Enable “Safe Boot with Networking” to isolate software conflicts.

ACPI/Driver Debugging

In Windows, run:

batch
powercfg /energy # Generates power report for ACPI errors
verifier /standard /driver *.sys # Checks driver stability
Review energy-report.html for USB/NVMe power errors.

rdxmaster

It happened to me last week as well with an EQ12 that had been running fine 24/7 for almost 2 years.
It had random reboots, sometimes even in the BIOS POST before the OS started.
I also tried memtest86+, but it rebooted with that as well.
I replaced the PSU and validated the power with a multimeter to be 19.4V under load when the reboot happens.
In the BIOS, I observed jumps from 0.7V to 1.0V in VCore, and my assumption now is that the VRM is faulty, which means this device is going to the garbage can, unfortunately.
It did give me good 2 years of running ESXi on it, so I have mixed feelings about our friendship ending.
This is my second Beelink dying this year, with my newer S12 pro memory stick faulting after a few months of use, and needing replacement so at this point I have to say I am frustrated with the NUC world and may be switching away.

domquark

Thanks for the replies.
Support:
Unfortunately, I cannot get Ubuntu to load properly. the time from transitioning between the BIOS to booting into an OS is getting less and less. One unit did manage to get to a desktop, but rebooted after a few seconds. The other is now rebooting as Windows or Ubuntu tries to load.
rdxmaster:
I think that I have the same issue. It would explain why it’s getting worse as the VRM degrades.
Thank you both for the information. It’s a shame that this has happened, as I love the size and CPU power of these units.