The article I linked also happened to say that the carbon nanotube transistors they came up with were also better than gallium arsenide ones!
Also, I’d like to point out that carbon nanotube transistors have existed for some time now. It’s just that today, they’re now better than silicon ones.
While we’re on the topic of non-silicon-based transistors, I had a professor once who was in phonon research and calculated the effectiveness of a phonon transistor (as opposed to the electron transistors we have now).
Just like how photons (from Greek prefix photo = light) are wave packets of light, phonons (from Greek prefix phono = sound) are wave packets of vibrations in solid-state matter. We’ve only come to know phonons in the field of solid-state physics, so their description is purely quantum mechanical.
High-amplitude phonons (large vibrations) are what we know as sound, while low-amplitude phonons (small vibrations) are what we know as heat.
We describe phonons only in solid-state physics; solid-state means the atoms are “locked” into position by electrostatic forces from other atoms. This happens when atoms are arranged into certain regular patterns, like crystal lattices for example.
My professor calculated how effective a phonon transistor would work. He imagined “wires” made of strands of atoms, through which current could flow. Since phonons have to flow through solid-state matter, you can imagine the strands to be connected via electrostatic springs.
By some mechanism (which I don’t remember lol), a potential could be made across one of the junctions, such that you could control when phonon current crosses. This gives the phonon transistor its switching action.
Unfortunately, phonon transistors turned out to be something like 4000 times less effective than electron transistors. As it turned out, you could make a computer without electricity (using only heat), but it’d be a terrible computer.