I would suggest two things:
- Download and run MemTest86+ and test your memory for a few hours (or day).
- While you are in Windows, run an application that can display (and better, one that can log) the temperatures of the CPU, Motherboard and the SSD(s). I personally recommend AIDA64. This is an app that costs money (worth it IMHO), but you can use it for a free 30-day trial beforehand. You can also run applications like SpeedFan and HWMonitor, but these applications don't truly understand the NUC boards and may display voltage and temperature sensors incorrectly (i.e. displaying non-existent temperature sensors and/or not displaying working sensors properly)...
Hope this helps,
Ok, but it happens once just after cold boot starting up the Nuc. So can it be a heat problem that soon.
And after the spontaneous reboot I have hit the F2 key to look at the BIOS and the temperature is normal within range.
Also after this reboot I ran Blender to render for hours 100% CPU use without a hassle. Or Crystal mark, flawlessly.
The nuc is on now all day, with no reboot whatsoever.
Used this tool a week a go: Breakin » Advanced Clustering Technologies
At work www.gigaserver.nl we build to order servers and use, amongst others, the breakin stress-test and diagnostics to 72 hour stress test.
The nuc runs the breakin test without any problems for hours.
Core 67° C, memory 68°C, mainboard 51°C, PCH 54° C
As I have said elsewhere, I believe that the issue is in the support for NVMe drives. Whether this is a hardware (board), firmware (BIOS), software (driver) or (particular) NVMe drive issue remains to be determined. I find it interesting that the problem seems to always be tied to the Samsung NVMe drives (there are zero reports of this issue occurring with SATA drives)...
Try to do a firmware update on your SSD, contact your system manufacturer in order to do the firmware update.
My guess would be the power adapter or SSD. Test the system with another one if it possible and test RAM again. It will be hard to isolate the issue if it happens randomly, reboots can be related to a bad capacitor as well.
Please let me know if this happens more often or if you need anything else.
Thanx for the advise,
The Samsung NVME storage has the latest firmware, memory seems to be fine.
Tested it with the tools we use for troubleshooting servers. The NUC gets "soaking hot'. The fan is spinning full speed.
When testing it. But the system remains stable. It will not reboot or freeze. Great job done by the designers ;-)
It looks like only to happen after a cold boot.
After 'cold' starting up the NUC and logging in to windows 10 pro after 15 to 20 minute the NUC restarts.
Then it stopt after rebooting at a black screen with white letters a message is saying:
BIOS has detected unsuccessful post attempt(s).
Possible causes include recent changes to BIOS
Performance Options or recent hardware change.
Press 'Y' to enter Setup or 'N' to cancel and attempt to boot with the previous settings.
Now you can Y enter the UEFI(BIOS) or just press N
So the solution I use now is to ALT F4 in windows to do a manual warm reboot.
After the second reboot there is no message screen as mentioned above and
after that the system is stable all day even under heavy load. it will not reboot again.
I also am experiencing the issue with a random reboot not long after powering up the NUC6i7KYK2 in the morning.
I never had the problem with this NUC the first several weeks I owned this......until the last couple of days after I upgraded the BIOS from 0034 to 0035. My sense is that there is something related to that BIOS version that is contributing to the problem.
Try a BIOS recovery; you can download the latest bios version here:
Even though you have the latest BIOS you can do the bios recovery with the same BIOS version 0035.
Please see BIOS Recovery Update Instructions for Intel® NUC
Very same experience here: unexpected shutdown 5 Minutes after cold boot even when doing nothing. I am running Windows 10 on a Samsung 950 Pro 512GB, but on a NUC6i5SYK.
- Upgraded to latest BIOS 044, also did BIOS recovery
- I ran memtest for 2 days without errors
- I ran Prime95 100% CPU during 4 weeks 7x24h without any hickup
As a workaround, I put the NUC into sleep. After resuming from sleep, I never had this issue.
Quite interesting statement from Scott, that these issues are with NVMe drives only. It looks more like a physical issue causing the NUC to shutdown, as Windows is not even able to present BSOD.
Done a recovery BIOS. "And till now no cold reboots anymore ;-P" So fingers crossed. Nope it still reboots :-(
Update: A BIOS update to KY0037 was the solution. To get rid of the cold reboots.
The procedure was a bit different than said in the instructions. And the jumper went broke.
If you remove the jumper and power back on and start the NUC, after first placing a@ USB thumb drive with the correct bios file in one of the USB ports.
The system keeps looping and looping starting up, shutting down.... So maybe the USB stick was in the wrong port. Or to big a Gigb. USB 2 or 3. Whatever.
Nope nothing seems to work. Even changing the bios file name in all CAPS did not help.
And when removing the jumper the metal part was left behind. Still connecting the pins. Hardly to see. So I had to pry the metal part off as well. And put it back together later.
After placing the repaired jumper back and powering up the NUC a screen popped up with options. The last line said hit F4 to do a BIOS recovery. So I did.
After recovery is finished power off and removed the power cord.... Power on again, push button started up the system. Did not go in to the UEFI/BIOS.
But linea recta into Windows 10. Powered off the system pull the plug from mains(electricity). Waited a while.
Then reverse order, logged to windows typing this reply. And experienced no reboot.
So tomorrow Dutch local time, I will try again. And let you know.
After 8 hours away, cold booted up the NUC. And guess what. It still reboots, same behavior.
So back to square one ;-(
A BIOS update to KY0037 finally was the solution. To get rid of the cold reboots.
Thank you all for the information provided on this post, orkey I hope you can have a better experience with the Intel® NUC from now on.
iPat is it possible for you to try different power adapter just to try to isolate the issue or you're not experiencing the issue anymore?
What is the rationale for doing a BIOS recovery if the system is running properly and if the BIOS installed without errors? Is the assumption that there is some errant bit flipped somewhere? Wouldn't that be detected by checksum verification when BIOS loads? I was under the impression that BIOS recovery was necessary only when it was completely broken and unable to load.