According to the discussion at the end of /thread/84871?start=31 this thread, the stick will assume it is headless if nothing is detected by the BIOS and not initialize the graphics for the stick.
This is causing an issue for me as my stick is attached to and powered by a TV. It appears the HDMI port of the TV is dead until ~7 seconds after the TV is powered on but the USB port immediately has power. This causes the stick to boot headless as the stick POSTs faster than the TV boots. The option to force the HDMI port to be active or the option to delay the BIOS's status check of the HDMI port would be fantastic.
I've tried most versions of the BIOS and they all see to exhibit the same behavior. Not that it matters, but this is a Win 8.1 stick upgraded to Win 10.
The stick will not initialize the HDMI port if no HDMI connection is found by the BIOS under USB power or the wall adapter.
The behavior makes sense (And is expected, according to the thread I linked above), this stick should be a low-power as possible - But having the ability to override a headless mode would allow me to move forward using this stick for a digital display project.
You can recreate my situation by powering on a stick, counting to 5, then plugging in the display...the graphics won't be initialized so the display remains blank. I'd like to override this hardware-level behavior as this is essentially how the HDMI port of the TVs I have tried (3 of the same model) seem to work.
Yes, that is correct is an expected behavior, note the picture below;
For further reference http://www.intel.com/support/motherboards/desktop/sb/CS-035558.htm Intel® Compute Stick — Issues When Using USB Ports for Power.
So is my feature request being ignored? I need to know if this will ever be possible or if I should move to another company's "compute stick" that will take a BIOS feature request.
Why you're bringing up power requirements is beyond me. Please take this question to an engineer or someone who understands the request. This has absolutely nothing to do with power requirements, as I clearly laid out in my original message.
I don't think this is a Compute Stick specific issue. As far as I know what Choong experienced at the end applies to all HDMI and DisplayPort connections since they are different from how DVI works.
You can easily see the difference on any Linux KMS driver with a display card and monitor with all three types of ports. For DVI, it will do modesetting successfully as long as the monitor is connected. For HDMI and DP, the monitor has to be connected to power and turned on, otherwise the kernel will not detect them and won't do modesetting for them (assuming you only have one physical connection at a boot).
But in Windows you shouldn't experience any serious problem because as long as the graphics driver is installed, it should do probing so you should be able to boot headlessly and connect to HDMI anytime you want. (But yeah this is OS/driver level, you probably can't press any key to UEFI setting and turn on the TV later and expect to see things)
I do not have Intel Compute Stick but the Lenovo one, almost exactly the same board but only with different UEFI (Shitty 32-bit Insyde). I can boot my stick headlessly and get video when I connect it to my TV later.
Thanks for the input, but I've tried several different things...along with @Amy_Intel confirming that when the HDMI isn't connected, it is not available for Windows to use... But knowing your Lenovo does allow you to connect the HDMI later is helpful -- except I understand the Lenovo doesn't have a "always on" option like the ICS so I'm still stuck without a perfect solution...
I'd prefer to scoop up a bunch of Win 8.1 versions of the ICS for this project, but I'll have to wait and see if Intel is interested working with the community to make their product better.
Don't know how much AMI would care about one's idea but apparently this is a UEFI issue and it's pretty much up to the UEFI vendor whether to improve things like this AFAIK.
Contacting Intel's vendor to push an update for Intel's product isn't going to happen. That isn't how hardware vendors work. Plus, I'm asking for a very common feature found in everything from Raspberry Pis to my Asus laptop.
I think you've got the right idea, though, going with Lenovo seems to be the right direction - I've figured a way around the always-on limitation so I think I'll just scoop a bunch of those up. I was just hopeful I would get an answer from Intel, and not a discussion of power consumption...
Well I feel the opposite way, especially when this doesn't look like a device-specific issue, but how AMI Aptio V generally handles headless boot.
If they do care about their users and consider this a bug, they are gonna fix it and let Intel know, then Intel will accept it happily (and lazily). After all I do think they know more about the UEFI itself. But surely you might end up getting a reply like "please contact Intel for this".
Anyway if you feel like you want to switch to Lenovo, make sure you try one before getting a bunch of it, coz its UEFI from Insyde seems glitchy to me as well and it's 32-bit only. Not to mention that Lenovo never released any UEFI updates so far and doesn't seem to have a proper product page for it .
Wow, does Intel's 'helpdesk' actually read and understand questions?
Minimally, some modern TV's have USB ports supplying 1A and even the unit not under load can power up with only 500mA. But that isn't even the question asked in this thread.
I'd worried more about corrupting the device with it tied into the USB power, but I guess ensuring it's powered down prior to turning off the TV will protect against that.
Thank your comments and feedback.
When addressing to the this post I approached it as a possible power issue, in fact this case was discussed with the forums team and we all had the same assumption, we even tried to run the test on several T.V's but as expected from the USB ports on televisions did not supply enough power, so we could not continue the test and we declare this as a power issue. Now that you have made clear what the issue is about, I will forward this case to the engineering team in order to get more information let them know about this request, as soon as we have an answer this thread will be updated.
I work for Intel Customer Support and I will be taking this request up to the engineering level but before I do that, would you please confirm that you are experiencing this behavior even when using latest BIOS https://downloadcenter.intel.com/download/25506/BIOS-Update-FCBYT10H-86A- Download BIOS Update [FCBYT10H.86A] ?
I am also assuming this issue report is for https://downloadcenter.intel.com/product/86612/Intel-Compute-Stick-STCK1A32WFC Intel® Compute Stick STCK1A32WFC only, at least for now.
Feel free to add more details to this post, I will get back to you by Wednesday of next week.
I can't say anything with respect to the mswindows side of things, but I can say that with the bios dated November 5th, I am able to boot the stick with the HDMI *unplugged*, then plug it in, and observe the desktop exactly as expected. This is under factory ubuntu. The only missing piece is that sound doesn't work if the stick powers up without the hdmi plugged in, but that's a driver issue which I believe I can solve with some "force it" options.
In other words, the current bios supports the functionality you are looking for.
I apologize for the delay in getting back to this thread.
We (Intel Customer Support) did and I personally did extensive testing and I was able to boot the Intel Compute Stick unplugged and plugging it back again didn't show any issue. Needless to say that in all cases I had the Intel Compute Stick connected to a power source different than the T.V.
I would recommend that anybody running into this issue contact http://www.intel.com/content/www/us/en/support/contact-support.html Intel Customer Support and open a Service Ticket. Please make sure that you make reference to this post and that the amount of units affected is clearly captured.
I am really sorry that you are facing this behavior, perhaps just a bad unit.