An eGPU dock that's actually SFF (name TBD)

[nox]

it doesn't necessarily need additional holes for a bracket, but it really depends on the bracket you use. you could even place four holes on the bottom with an additional L shaped bracket for the vesa. but this all starts relying on what vesa bracket you plan to use and which mount. i guess standards need looking at.

also back in the old days, computer PSU's used to have a 240v passthrough outlet - i don't know why they stopped, but if it wasn't some EU legislation, it might be worth looking at for the monitor too - one less cable.

 

[iFreilicht]

also back in the old days, computer PSU's used to have a 240v passthrough outlet - i don't know why they stopped, but if it wasn't some EU legislation, it might be worth looking at for the monitor too - one less cable.

It seems like there simply was no requirement for it anymore when the ATX standard for PSUs was drawn out. From what I've gathered, AT PSUs had this outlet to power the monitor, so that powering down the PC cut the monitor from power as well to save power. Since then, monitors became a little smarter and will go into a sleep mode that draw little power, and since December 2012, a device that's off/in standby is required to use less than 0.5W of power.

An additional factor is that PCs powered by AT PSUs had a power switch at the front of the case that actually cut the mains power directly, so all power for both monitor and PC was flowing through that switch (which isn't a sustainable way of doing things when the PC can draw more than 1000W of power). This meant that the monitor output was easy to wire up, but with the ATX way of using +5VSB and the #PS_ON signal to turn the PC on softly, the PSU would have to include two relays to switch the monitor power on and off, which take up a lot of space.

Of course, that doesn't mean that one couldn't include an AC passthrough in any new devices. Maybe there are safety concerns, but I couldn't find anything on that.

 

[Arboreal]

It seems like there simply was no requirement for it anymore when the ATX standard for PSUs was drawn out. From what I've gathered, AT PSUs had this outlet to power the monitor, so that powering down the PC cut the monitor from power as well to save power. Since then, monitors became a little smarter and will go into a sleep mode that draw little power, and since December 2012, a device that's off/in standby is required to use less than 0.5W of power.

An additional factor is that PCs powered by AT PSUs had a power switch at the front of the case that actually cut the mains power directly, so all power for both monitor and PC was flowing through that switch (which isn't a sustainable way of doing things when the PC can draw more than 1000W of power). This meant that the monitor output was easy to wire up, but with the ATX way of using +5VSB and the #PS_ON signal to turn the PC on softly, the PSU would have to include two relays to switch the monitor power on and off, which take up a lot of space.

Of course, that doesn't mean that one couldn't include an AC pass through in any new devices. Maybe there are safety concerns, but I couldn't find anything on that.

Yes, I remember AT PSUs in my first few PCs; you could get a pass through cable for the CRT monitor as mentioned.
I also remember the fateful day that I got a shock from the front switch in an AT PC. I was working on it , still plugged in...and the insulation boot had slipped off one of the spade terminals on the front power switch! I did get a bit of a belt, but not enough to throw me.
Thank goodness for ATX; it meant that PCs could catch up with Macs and do software shutdowns at last

 

[nox]

'It's now safe to turn off your computer.' wasn't exactly apt then?

nox

 

[iFreilicht]

'It's now safe to turn off your computer.'

Now I finally realised why that was necessary in the first place. Wow, those were weird times.

 

[nox]

Now I finally realised why that was necessary in the first place. Wow, those were weird times.

yeah, those AT (Animal Transporting) PSU computers had to retrain the hamsters to check the switch sheathing too before displaying that message on the screen. They had a much easier time back in the days of DOS. I heard they actually marched on Microsoft's head office over the additional work with no benefis.

nox

 

[TTheuns]

Can I just add that VESA mounts on the back would probably not be an issue aesthetically. I mean, I haven't seen the 'back' of my PC for weeks.

 

[jeshikat]

Good points.

I'll be doing a live stream this Saturday and I'll play around with the VESA mounting ideas.

 

[jeshikat]

I'll be streaming at 3:00pm Central (click here to see in your local time zone).

 

[jeshikat]

And we're live! https://beam.pro/SFFNetwork

 

[Runamok81]

I missed the live stream. Any new decisions / additions? VESA mounts?

 

[jeshikat]

Added VESA mount holes on the back, and I added a cutout for the C14 inlet for the AC-DC since I have completely forgotten about it until now.

 

[pyxls]

@Aibohphobia Will there be a stream today?

 

[jeshikat]

Sorry, no stream today.

 

[jeshikat]

Archive for eGPU 4:


I'll get 5 and 6 sometime next week.

 

[Runamok81]

Thanks for the archive!

 

[nox]

another option is to make it smaller and more vesa shaped by making it compatible with mxm GPU's instead... that's if they actually become sensibly priced (after Asrocks efforts recently.)

Nox

 

[EdZ]

That would either need a custom Thunderbolt host board, or a rare-as-hens-teeth PCIe card-edge-to-MXM breakout board that isn't huge.

 

[nox]

wow that thing needs to go in the space inefficiency thread!

 

[Runamok81]

That would either need a custom Thunderbolt host board, or a rare-as-hens-teeth PCIe card-edge-to-MXM breakout board that isn't huge.

I now know that this exists. Thank you.