What happens:
I am experiencing some issues with my brand new Skull Canyon NUC. This manifests in either a full reboot with zero warning, or the display freezing and the last couple of milliseconds of audio in the buffer endlessly repeating until the system is manually reset.
Some browsing around these boards indicates that there are a lot of users with the same configuration that are experiencing the same problems, however for the sake of completeness here is my setup:
- Skull Canyon NUC (NUC6i7KYK2)
- 2x 256gb Samsung 950 Pro SSDs (configured in RAID 0) - firmware version on both drives is 1B0QBXX7
- 1x Corsair 16gb DDR4-2133 SODIMM (part # CMSO16GX4M1A2133C1) installed in the first SODIMM slot
- Windows 10 Pro 64-bit
Upon bringing the system back up, there are no bugcheck logs or anything else in the Windows event viewer to indicate that anything went wrong, other than a generic error noting that "The system has rebooted without cleanly shutting down first. This error could be caused if the system stopped responding, crashed, or lost power unexpectedly". From this I can infer that whatever is causing the problem is fairly low-level (thermal overheating, RAM or something else outside of Windows' control).
Diagnostic steps I have taken so far:
- Updated system BIOS to the latest version (0042).
- Re-installed graphics drivers (also tried a clean install of the latest Beta driver).
- Adjusted CPU fan curves to kick in at a lower threshold.
One step I haven't been able to take is to update the firmware on my 950 Pros. This is because Samsung Magician is unable to see past the Intel RST array and thus can't access the drives directly, at least from Windows. If anyone knows of how to work around this (e.g. via a bootable disk image), please let me know.
I'm also going to run memtest86 overnight and see if anything looks iffy, but based on reports from other users on these boards I have a feeling that this is probably not the cause of the issue. Will report back with the results when I have them in any case.
(Edit: ran memtest overnight, no problems found.)
How to reproduce:
This seems to happen fairly often whilst gaming, although I have also seen it happen once whilst just sitting at the Windows desktop. The games I've been using to test include Grand Theft Auto 5, and I've also seen the same problem with Half-Life 2 (which is a bit bizarre given that it's a ~12 year old game and really should not be taxing to run on modern hardware). The problem usually occurs after about 5-10 mins of gameplay (although adjusting the fan curves seems to have extended this to anywhere from 30 minutes to an hour before symptoms manifest).
With that said, I've also used the system to watch multiple-hour long movies without any problems. General desktop usage (light web browsing, word processing etc) also does not seem to be a problem.
So, what's the deal here? Is this a thermal issue, part compatibility issue, or a firmware/driver/BIOS issue?