Ran all tests (CPU, Memory) and all passed without any errors.
It's with a new NUC8 model : NUC8i5BEH, Crucial 16GB Kit (2 x 8GB) DDR4-2400 SODIMM, Intel® SSD 660p Series (512GB, M.2 80mm PCIe 3.0 x4, 3D2, QLC)
Suddenly when system is Idle or working in windows in basic usage (internet/office etc) the computer freezes.
When it freezes i don't have access to anything. Don’t get any BSOD and Cannot ctrl+alt+del / ctrl+shift+esc.
Cannot move the mouse cursor and cannot type anything in keyboard.
The only solution is the force shutdown by a long press on the power button and to power on the nuc again.
I've disabled all power options (sleep/hibernate/screen etc).
1 more issue that might be related - i have a 144Hz screen so I am using the 144Hz option (is it related).
There is no event in event viewer and no log.
Latest drivers of the NUC on a clean installation of win10 pro.
Hi @NLaor (again)
Ran the quick scan and it passed successfuly.
I will try to update the FW and will update you...
For now i've already updated to latest drivers using https://www.intel.com/content/www/us/en/support/intel-driver-support-assistant.html (Intel® Driver & Support Assistant)
and i have the green message : Your Intel® software is up to date.
As I said in my responses in the other conversation, I believe that we are staring at a graphics driver issue. I would like you to install the latest DCH driver release (available here: https://downloadcenter.intel.com/download/28646/Intel-Graphics-Windows-10-DCH-Drivers) and see if this makes a difference.
Hope this helps,
So After 2 days and 8 hours or so the computer froze again.
I've turned it off and on 2day and when typing the reply to this forum it froze again (after 2 hours or so).
Do you have any idea what else should I do ?
Should I try to see if it freezes with only one memory card SODIMM ?
The Windows diagnostic doesn't do much. I would recommend using a tool like MemTest86+ (it's freeware) and let it run for a couple of hours minimum (I usually let it run for 24 hours as a burn-in test for new NUCs and new memory). I would do this with both SODIMMs in the system.
Yes, I do recommend trying with one SODIMM instead of two. I am torn as to which I would prioritize. Seems like running this test is easier to do initially. Remember though that there are 4 permutations that you would need to try: SODIMM A in Upper Slot, SODIMM A in Lower Slot, SODIMM B in Upper Slot and SODIMM B in Lower slot.
Hope this helps,
I have another NUC in my house.
Model Intel® NUC Kit NUC6i3SYH
It has 16GB RAM as well but instead of 2400 it has Crucial 16GB Kit (8GBx2) DDR4-2133 SODIMM
The nuc6 works for almost 3 years without any issues...
Do you suggest to switch the SODIMM ? I didn't ruin the MEMTEST86 yet (will try to run hopefully during the weekend).
Let me know what you think.
Also, to complement Leon's comments, testing with only one SODIMM (at a time) is important and testing now (after a month of break-in time) is better than your original testing.
If you mean use these SODIMMs from the SY NUC to test the BE NUC, yes, that's a good idea. If it works, you could then try using the 2400 SODIMMs in the SY NUC and see if this works (or identifies an issue with them).
Did you get to the bottom of this ? I've got an NUC8i7BEH purchased in Dec 2018. It was running fine until about a month ago when I upgraded to Windows x64 1903 - before that I'd run the Intel Driver and Support Assistant and it upgraded my BIOS from 51 --> 71. Since then the machine has been completely unstable. Randomly freezing, sometimes rebooting sometimes not.
My system: NUC8i7BEH, 16Gb Corsair Vengeance Memory, Intel 256Gb M.2 760p SSD
On the occassional times that it does reboot, I'm getting DRIVER_IRQL_NOT_LESS_OR_EQUAL and others PAGE_FAULT_IN_NON_PAGED_AREA in the dump files. Sounded like a driver issue to me.
I've tried everything (reverting to Windows x64 1809, various GFX driver versions - though I'm am unable to revert the BIOS to 51 using any of the techniques on the Intel website. I think they've locked 71 to prevent reverting to one of the bad BIOS's released between 51-71 which were withdrawn.)
I have a sure fire way of making my machine stable again - I've had to change the "Intel Iris Plus 655" drivers to "Microsoft Basic Display Adapter", though doing so means I lose HDMI audio and QuickSync. I'm extremely suspicious of this driver (I've tried the version from December last year which ran stable for months, the "recommended" version from IDSA, and the latest 6912. The machine will hang if simply left idling at the desktop after a random amount of time, up to several hours. It's very similar to the problems reported here https://ipcamtalk.com/wiki/memory-leak-quick-sync-hardware-acceleration/ I've been running Plex Media Server which has QuickSync accelerated hardware transcoding and it seems to crash quicker if I use this feature. It still crashes if I do not use Plex, simply boot, login to desktop and leave the monitor on (disable sleep, monitor timeout etc)
I guess it might be a hardware problem in the GPU part of the CPU (though as I mentioned changing the driver to "basic" works around the problem). The only other thing I can think off is to try a SATA SSD rather than the M2, as there were reported problems with other brands such as Samsung. I've got the lastest Intel SSD Toolbox and ensured I've got latest firmware on it.
I'll need to open a new case with Intel as "the support case associated with this question has been closed. Discussions may continue but if you need help from an Intel representative, please create a new question."