I made a previous comment and suggested a possible solution on a previous thread, but because the Intel admins have "Closed" that case with a non-solution, it's buried in the results on this poorly designed community site. I'm just going to copy and paste my solution for the sake of those experiencing this issue and wanting some kind of easily discoverable resolution.
My comment with possible solution:
I, too, have this problem. I use two external WD Elements drives and one Seagate. All have their own power source. However, all three plugged into the rear USB 3.0 ports on my NUC8I7HVK cause boot to fail and result in the POST message. It's somewhat ridiculous that the Intel admins on this board cite compatibility when the external drive business is entirely dominated by WD and Seagate. Designing a computer with incompatibility issues with, essentially, the entire external HD market is Intel's issue, not user error.
My theory is that the NUCs have a BIOS bug that believes these drives aren't self-powering and, thus, tries to feed them the requisite power but shuts down when it proves to be beyond the USB 3.0 spec for such large spin drives. One drive plugged in doesn't trigger this message. Two drives usually doesn't trigger this bug. Three drives plugged in always triggers this bug. So it's an issue of aggregate power draw from all of the ports. Another poster has encountered this problem:
His solution was to go into visual BIOS and set the rear ports to "No Detect" during boot. It's an old post from the 5 series of NUCs, but this problem clearly still persists. I've implemented this solution (the drives will still load and appear once windows boots up). Once I've gone through enough bootcycles to say something conclusive, I'll update this post.
I regret buying this NUC. So many issues. So many headaches. Intel can expect a bill from any drive failures this thing manages to cause
First of all, the BIOS does not have anything to do with power. The USB circuit provides the requisite power level independently. The only control the BIOS has is to turn off the power to the port completely during certain states. The NUC will boot just fine with a single disk installed; it is when a second or third is installed that the boot problem occurs. Further, I have shown that connecting one drive from each (Seagate, WD) vendor does not have any issues. The issue has something to do with the use of multiple drives from the same vendor.
All that said, I too agree that this is a (unfortunately longstanding) BIOS bug that is not isolated to this one NUC model, I also agree that any discussion of compatibility is ludicrous and any issues of drive availability plain silly; these are the standard drives used almost exclusively by the market.
There are three states for USB ports in BIOS: Enable, Disable, and No Detect. What I didn't clarify in my original post (but others have) is that multiple drives connected to rear ports with the Enable state hangs startup until the drives spin up to some degree. I can't be sure, but this might also contribute to the boot fails. The hang time is certainly not what you would expect given this NUC's capabilities and an OS on an NVME stick; the windows boot screen should appear in under 5 seconds from power up.
I'm unable to test this solution of setting the rear ports to No Detect for a few weeks because of the holidays. I would highly appreciate it if you or others with access to NUC8s and multiple external drives can test this for the community and we can get some various data points.
First of all, BIOS boot time has very little to do with the performance of the processor. Secondly, the speed of the boot drive only has a bearing on the time after POST completes. It is factors like USB readiness - presence of USB hubs (beyond those in the PCH), connected devices, etc. - that have the most bearing on POST times. You are absolutely correct; if USB-based drives are not ready, then the BIOS is going to have to wait for them (this is a spec-level requirement). If you have drives that are semi-permanently connected and you are unwilling to wait for them to become ready, then you should disable detection on these ports (that's one of the reasons why the feature is there).
I personally do not care about boot time; systems shouldn't get booted that often so the time taken is not a consideration (and no, I am not interested in debating this). The BIOS having to wait would normally be a non-issue for me but, considering this hang issue and the fact that I consider this a BIOS bug, these waits represent a data point that might help with the tracing of this bug. Regardless, it is time for this issue to be investigated by the NUC BIOS team. I await their analysis...
Thank you for posting on the Intel® communities.
We appreciate all the details you provided regarding this behavior with the boot failure when using self-powered external hard drives. We will check this further and as soon as we have more details, we will be posting back in this thread.
Also, could you please confirm the BIOS version running on the Intel® NUC Kit NUC8i7HVK?
Intel Customer Support Technician
A Contingent Worker at Intel
I have the latest BIOS version:
BIOS Update HNKBLi70.86A
Latest version: 0059
I had the latest BIOS when I acquired the third HD. All of these recent problems occurred while on the latest BIOS. All of my system's other firmware and drivers are updated to the latest available through IDSA.
Thank you for your response and the details regarding the BIOS version.
We will check this behavior further and as soon as we have more details available we will be posting back in this thread.
Intel Customer Support Technician
A Contingent Worker at Intel
Can you please provide me with the Brand/Model of the external drives that you have tested?
I just tried with my NUC8i7HVK running the latest BIOS (BIOS set to defaults - USB ports set to enable) and Windows 10 with an external WD* drive model WD 4TB My Passport USB 3.0 (WDBYFT0040BBK) and I had no issues.
Please reread my posts. The point is that multiple drives plugged in cause the issue, not a single drive. As i stated earlier, I have three drives - One seagate and two western digital. The seagate is STGY8000400. The two WDs are WDBWLG0080HBK-NESN. These are all also large capacity drives that require their own power source. The passport WD you tested is powered entirely through the USB port. If the intel engineers want to address this problem, please acquire a few large external drives from both manufacturers.
I have tried with two bus-powered drives (WD+Seagate) without seeing issue. Must be tied to slower, self-powered media.
We're working to reproduce this issue here so that engineering can debug it. They have to see it in order to fix it. :)
Got a couple questions for you.
lois.hill - I privately messaged you my S/A number.
I am not using any of the three external HDs as boot drives. They have no OS installed on them. They are just there for bulk storage.
Update: I'm back at my unit after a few weeks away and have gone through 3 or 4 boot cycles dealing with updates and such. Since I've switched to "No Detect", I haven't had any boot failures and POST messages. Boot time still takes a time premium though with the drives attached.