Morning and, firstly, thanks for the reply.
Unfortunately, the case here isn't USB device related, and it just so happens I've got an example from this morning.
I use a machine of the same type, and I have many USB devices plugged in.
From time to time, windows will blue screen, every 3 or 4 days (my machine is constantly left on), and when it reboots, it exhibits the behaviour I outlined in my original post.
My machine is strapped to the back of my monitor with the VESA mounting plate, is never knocked/jolted, and so nothing is coming loose. I then get this screenshot.
If I press any key, it just cycles around the same message (hence so many entries on the screenshot)
However, rather than power down at this point, I pressed CTRL-ALT-DEL and went into the BIOS, it was there that I've now noticed the drive doesn't appear listed as a boot device (see image 1 below), and I use the latest BIOS, v35 (image 2 below), and I've also set it to not even attempt boots from anything but the hard drive itself (image 3 below)
For good measure, I removed all USB devices in this instance (even though out in our branches, there will be NO USB devices plugged in), and it made no difference.
So, the ONLY way to resolve this is to pull the power to the NUC (the mains lead) and plug back in. Magically, Windows then boots up, but it works, and the BIOS shows the hard drive as a boot device. So, this is clearly not a windows issue, and I've got another of the i5 boxes with a different model hard drive, doing exactly the same thing, so it's not the one I'm using on this rollout.
Can anybody shed any light on this. There is obviously an issue here, and I could do with some answers
I've had a word with our suppliers and they came back with this...can anybody from Intel confirm or deny this? Has anybody else been told this?
Doing some quick research it looks like it's a hardware fault that drops the SATA connection at random intervals. Hopefully Intel will release a fix for this soon.
I looked for similar problems in all threads updated since 2/1. I've highlighted a few that reference a known hardware issue, though it's not clear if the symptom would be a RANDOM failure to boot, or if it would result in a more consistent failure (like never detecting the drive in the first place).
N2820FYKH0 ("flashed to the new bios") - https://communities.intel.com/thread/59908
D54250WYKH2 - https://communities.intel.com/thread/54719 <-- This one was originally created in August; Intel linked to another thread where the user had to have their NUC replaced.
- 4 users impacted
- D54250WYKH2 (BIOS unknown)
- D54250WYKH (BIOS 0021 and 0027)
- D34010WYKH (BIOS 0028)
- D34010wykh (BIOS "latest as of 12/29/14")
- Link in here to a PDF where user says "hardware rev 104 and later had a change to fix this issue" (on D54250WYKH)
D54250WYKH ("most recent version BIOS 0042") - https://communities.intel.com/thread/60193
- That BIOS rev doesn't exist.
D54250WYK (BIOS 0022) - https://communities.intel.com/thread/47495 <-- Thread from Dec 2013
- BIOS 0021 seems to be stable
- 2 additional users impacted
- 1 (hw/BIOS rev unknown)
D54250wyk (BIOS unknown) - https://communities.intel.com/thread/60332
D54250WYKH https://communities.intel.com/thread/57620 <-- Thread started in December
- User claims that the Intel SATA cable in the NUC is "non-standard".
- Intel sent updated daughter card, which apparently has the fix referenced in the PDF mentioned above.
- Multiple users say this hardware update has fixed the issue
D54250WYKH - https://communities.intel.com/thread/60157
And then in this thread:
D54250WYKH (BIOS 0026, 0033, 0035)
Many thanks for that information.
I've had a look through and of it all and I'm particularly interested in the PDF linked to the hardware revision (and the updated daughter card)
My hardware revision is H26247-102 - this I'm making the assumption I've also got a potential hardware issue because it's earlier than the 103 revision stated in the PDF notification. I've let my supplier know and I'm going to see what they come back with as I don't fancy explaining this to the Intel live chat bods myself (they should be reading this anyway!)
In the meantime, I've upped the Minimum Duty Cycle setting (found under 'Cooling' in visual BIOS) to 50% and 60% across some different hardware, and now playing a waiting game for the random crash, just to see if I can establish pattern, but I'm not hopeful
Can anybody from Intel, on this forum, actually grow a pair and tell me if I've now got 400 duff units out there that need hardware replacement?
My batch details are as follows
Product Code: BOXD54250WYKH3
Version #: H26247-102
Manufacture Date: 03/07/14
You should contact Intel for a daughterboard swap and see if it solves the problem.
If it does, then you'll be able to go back to your vendor and tell them what you need (the 104 HW rev).
If it doesn't, then you need to get a ticket filed with Intel support or just return the 400 devices and go with something else. (I think you should at least give phone/chat/email support a try before going that route)
We recommend the following steps in order to address the issue described in this community thread:
• Install the Intel Rapid Storage Technology Driver 184.108.40.2067 (or later).
• The SATA daughter card integrated into the Intel® NUC D54250WYKH and D34010WYKH has been updated please contact Intel Customer Support and ask for a card replacement. See following URL for further details on how to contact Intel Customer Support: http://www.intel.com/content/www/us/en/company-overview/contact-us.html
• Replacement SATA daughter cards (part number H30274-005) are available for customers who encounter this problem and do not use the Intel Rapid Storage Technology Driver 220.127.116.117 (or later).
It seems Intel have now acknowledge that the PCN in question does cover my boxes, which are a mixture of 102/103 hardware revisions.
So in terms of a solution, my supplier is now managing it directly with Intel and instigating a rolling program of replacements.
I appear to have had some success with changes made to the Cooling | Minimum Duty Cycle setting in the BIOS, and upping this to 60% on some machines have meant they haven't crashed in the last 6 or 7 days, but it remains to be seen if that is sustainable or whether it just goes longer without crashing. I've also got a batch shipped @ 50% but they're not in yet.
Unfortunately, I can't stop the rollout because we can live with the odd reboot for what we get overall in terms of branch advertising, so it's a pain to tidy up later, but careful planning should alleviate that somewhat
Good luck getting it sorted though, I'd definitely contact Intel to see if you're affected, they were really good on the phone (I got the UK office in Swindon)
To add to the group-
Yesterday I changed the "Hard Disk Pre-Delay" settings on one of my problem NUC's to several different values greater than 0 but the box would not boot at any setting.
In order to help isolate the offending problem, I ordered up a mSATA drive, grabbed a back up of the device (Acronis) pulled out the SATA SSD then restored the machine using the mSATA drive. After that was done the machine rebooted several times without issue (the NUC has probably rebooted 6-10 times since the drive swap and I have not had one failure to boot).
I am going to give it a week before I commit to calling it a fix, but so far so good.
I realize this may not be a fix for addy and his 400 machines, but I hope it helps confirm the failure of the daughter card hardware.