According this Intel info I mean: Intel Product Change Notification for Skylake NUCs . As we can see, there are 3 important infos:
1, new BIOS release 0039 (it's on download page yet)
2, changing capacitors (..as I still said before )
3, old and new Part Numbers for 6iX NUCs
I upgraded to 0039. No issues yet...I'll monitor temperatures for sure...
I have upgraded BIOS to 0039. All looks good. But I left the capacitors unchanged
Anybody with WHEA errors install 0039 and the errors have stopped or at least gone down to a minimum?
Anyone got a link to the new BIOS firmware? I cannot find it!
I have updated My NUC6i5SYH to BIOS 0039 and apply WA (disable C2) indicated in document "FAQ for SY.pdf". It works fine for both Windows 10 and Linux!!
My configuration is the followings:
- Memory: CFD panram W4N2133PS-8G x 2 (Taiwan ram chip vender's module. not familiar one) -- 16 RAM chips are mounted per module
- SSD1: SanDisk Z400s SD8SNAT 256GB (M.2 SATA) -- for Windows
- SSD2: TOSHIBA THNSNJ256GCSU (2.5 inch SATA) -- for Linux
- OS1: Windows 10 Pro 64bit build 10586.164 (Clean installed)
- OS2: CentOS 7
- USB device: Keyboard and mouce
- Display: Acer V206HQL HDMI via DisplayPort - HDMI conversion cable
Before update to BIOS 0039, Windows 10 crushed with various type of BSOD (but I have't experienced WHEA) in about 30 minute after login to desktop. Now it is up and running for 6 hours including run of Direct-X benchmark.
As for Linux (CentOS 7), I had kernel crush on SSH benchmark test using bonnie++/fio. After BIOS 0039, it works fine with default kernel of CentOS 7 (kernel 3.10) . As I am using NUC for server purpose, I haven't installed X-Window/Desktop, so kernel 3.10 is fine for me (no need to install 4.3 or later which include improved support of Skylake).
I have struggled with unstable linux kernel about a mouth since I bought NUC6i5SYH. Finally I've got stable machine.
However, I want update of my NCU same as PCN-11478 (I guess that is improvement of power line of memory bus).
I have updated My NUC6i5SYK to BIOS 0039 and disabled C2. My value temps are 50-60 C when watching video. At least 10 Cs lower.
However I do get max spikes up to 74Cs. I don't know what triggers the spikes.
I've also just updated to 039 from 036, but l have been lucky to so far not had any WHEA errors since purchase .... but I have only had mine for 9 days now.... have to admit I am a bit afraid to use the NUC. I originally updated to 036 as soon as I took it out of the box. I also have the unit configured for high performance, and I have not messed with C2 states.
Based on what Intel has already published, my NUC falls in to the category of one that "could" be affected. That having been said, I am not having any issues, not yet anyway. so according to cvare the change is primarily to deal with certain types of DDR4 memory ??
Is there a way to know if a particular kind of memory is likely to cause issues? In my case my memory is 2 sticks of: G.SKILL Ripjaws Series 16GB DDR4 2400MHz CL16 1.2V SODIMMs - F4-2400C16S-16GRS
I didn't mess with the bios, letting it run the memory at 2133Mhz ....
Is there a list of RAM modules / part numbers for memory that there is evidence of issues ? a list of those parts would certainly be handy ....
I still getting WHEA BSOD error message after update BIOS to 0039 and disable OS C2 report. I just want to clarify the problem is related to memory compatible issue or motherboard hardware fault.
Everyone can get WHEA error. It doesn't matter which RAM we are using. BIOS 0039 doesn't solve WHEA. If you receive new NUC with replaced capacitors, you still can get WHEA error. New capacitors only assure that all RAM modules should work in NUC correctly. But WHEA error is different thing which isn't related to RAM. It's different problem and Intel is working on it.
...and so If I have intended right, because this new revision of the board don't fix the WHEA errors and Intel are still investigating what can cause these errors, they made a new revision of hardware for substantial nothing that is related to the problem?
It is like saying: "we still don't know what cause this WHEA errors that many users have, however for now we decided to change some production parts on the board to ensure better compatibility with ram, but this isn't related and don't resolve the problem that users complain about....
So even after the change, the product remains still bugged, and users will continue to receive units that don't work right even after April.
Or are they sure to fix lately these WHEA errors through a software update? and how it is possible to know this if they still don't know what could be the cause?
Also when Intel have said that new caps assure that all RAM modules should ok, It is implicit that now not all RAM modules work correctly in the nuc's. What ram? (I tried to ask again what these "early DDR4" modules means, because for me it means nothing)
And so more important what are errors caused by this fact?
For the test I enabled C2 with BIOS 0039, and I got Windows 10 freeze when I launch Direct-X benchmark software.
So, it looks like disable C2 is mandatory for my environment. After disabling C2, CPU temperature of idle condition rose about 2C.
My idle temp is the same after disabling C2 as before - I'm watching movie (right now it's paused) and writing this and temp is 37.C