I'm coming from a place of complete ignorance here, so take my question as genuine and not trying to imply that this _should_ be an easy problem. But what exactly is it that makes it so difficult to have a KVM that lets me connect two computers to two high definition (2k in my case) monitors, along with some basic USB peripherals and audio components and switch between them? Every single device I've found has had some drawbacks like not supporting high framerates (144hz), not supporting Mac/Linux/Windows, only supporting audio output and not a microphone, not supporting thunderbolt or only supporting low resolutions.
Is it just that there's no market for it and that the cost of it would just be too high? If money was not an issue, would there still be technical reasons that this is impossible?
Because those signals are really high-speed, and the protocols are really complicated.
Doing the equivalent of "yank cable from PC 1, plug cable in PC 2" is just about doable at a reasonable price point. Anything more complicated either requires a bunch of expensive hard-to-source dedicated chips, or a bunch of hard-to-implement software solutions. Especially stuff like reliable keyboard-controlled switching or USB-C laptop connectivity is a nightmare.
In practice this means you either have to give up on features, or let the price balloon to unacceptable levels.
Audio mixing would be a useful feature; there could a free-spinning volume slider or knob for each audio input, and an option to focus/isolate only the focused AV source
I'm coming from a place of complete ignorance here, so take my question as genuine and not trying to imply that this _should_ be an easy problem. But what exactly is it that makes it so difficult to have a KVM that lets me connect two computers to two high definition (2k in my case) monitors, along with some basic USB peripherals and audio components and switch between them? Every single device I've found has had some drawbacks like not supporting high framerates (144hz), not supporting Mac/Linux/Windows, only supporting audio output and not a microphone, not supporting thunderbolt or only supporting low resolutions.
Is it just that there's no market for it and that the cost of it would just be too high? If money was not an issue, would there still be technical reasons that this is impossible?
Because those signals are really high-speed, and the protocols are really complicated.
Doing the equivalent of "yank cable from PC 1, plug cable in PC 2" is just about doable at a reasonable price point. Anything more complicated either requires a bunch of expensive hard-to-source dedicated chips, or a bunch of hard-to-implement software solutions. Especially stuff like reliable keyboard-controlled switching or USB-C laptop connectivity is a nightmare.
In practice this means you either have to give up on features, or let the price balloon to unacceptable levels.
There is no source, nor even a way to order one or find out more. It was abandoned 4 years ago.
They are under separate repoes by the same person.
Audio mixing would be a useful feature; there could a free-spinning volume slider or knob for each audio input, and an option to focus/isolate only the focused AV source
Re: pikvm a "DIY IP-KVM Based on Raspberry Pi", DB9 & RS-232, AMT, DASH: https://news.ycombinator.com/item?id=38062923#38065133