Well CPU's are made of transistors. So one could ask if it is possible to make a transistor out of a superconductor.

The answer is yes. It is possible to quench a superconductor with an external magnetic field for example and make it an insulator. These things actually do exist but i forgot the name.

But switching that field still requires energy. So while it is possible to build such a CPU there would be still energy required to do any calculations.

Today's CPU's don't heat up because of internal resistance (mostly) but because there is current flowing due to the field effect transistors needing their gate voltages being raised and lowered.

Lowering that voltage means the energy must be vented somewhere which results in heat. That would not change if any part is superconducting.

Maybe watch this video about entropy. Its not about where energy flows, its about hoe entropy can be used to calculate.

https://youtu.be/DxL2HoqLbyA

>Where did the energy go then?

Through the circuit. Heat is energy loss.

The heat comes from the resistance. No resistance, no heat. To be more specific, the heat is from the kinetic energy of electrons bumping into the conductive material; no collision no heat.

Lets make a simple Gedankenspiel with only a wire made from LK-99 and a bunch of electrons sitting on the left end of the wire.

After a moment, the electrons will be spread out evenly through the wire, right?

So after a moment, the energy that was stored in the system (When all electrons were on the left) is lost.

If no heat is emitted, where did the energy go?

Current in a conductor does not involve the potential energy of charges (electrons) being concentrated. This should be clear by the fact that currents cause magnetic fields, but do not give the conductive wire a charge gradient (or an electric dipole moment).

That's not how electric current works. It's more like a wave in an ocean. A wave may travel from Hawaii to California but no actual water molecule moves that far (or even very much at all). Electric current is a wave of excitation of electrons, which is what moves at the speed of light. The actual drift of individual electrons in the current is much, much slower (and IIRC in a superconductor is zero).

Veritasium has a cool video on this:

https://www.youtube.com/watch?v=bHIhgxav9LY&pp=ygUVdmVyaXRhc...

Interesting video, So it is sad that they did such a terrible job explaining why the the answer is 1/c.

It makes it sound like the bulb is getting full power at 1/c, when it wont. As I understand from further reading, the power will build up over time, and that time depends highly on the shape of your circuit.

Why would the electrons spread evenly?

They move over to the other side and back and forth resulting in an oscillator.

That causes all your energy being radiated into the surrounding via radio waves.

That is how your microwave oven works but instead of a superconductor it uses a magnetron with free flowing electrons.

Nowhere, it's still there - a loop of superconducting wire would also be a perfect energy storage because the electrons would just keep going around without any loss(it would be an extremely small capacity storage though, superconductors lose their superconductivity at high enough amperage which is basically what introducting more energy into the system would be).

I suspect that a circular trajectory implies acceleration and when a charged particle accelerates, it radiates away energy.

Superconducting circuits can still have AC loss. You’re basically describing discharging a capacitor through a superconductor. If there’s no loss anywhere, it will be be an LC oscillator. If there is loss, the energy will dissipate.

I guess calculations are done using semiconductors. So you could perhaps get energy to a transistor without heat, but the transistor itself is not a superconductor and will still emit heat.

(Edit - the question is now different so my comment is stale)

My understanding is power is expended when current flows through a metal with resistance, and that loss is in the form of heat. The lower the resistance, the lower the loss and therefore lower the heat

But you do lose energy as current flows.

When you start with 10 elecrons on the left of the wire and none on the right and end up with 5 on both sides, you lost all energy that was stored in the system.

The mind-blowing thing is that electrons, as far as I understand it, don't actually flow through wires at all. If they did AC power would be pointless

Well they do but very slowly. With AC they just wiggle left and right.

This is some kind of a simplified view of how a battery works. It is not how an electric field functions.

> How much of the heat a CPU expells is inevitable?

There's a lower theoretical limit: https://en.wikipedia.org/wiki/Landauer%27s_principle

> Could (non-reversible) calculations be done without creating any heat?

Thermodynamics says no.

> When energy flows from one and of a wire made from a superconductor to the other end, then no heat is produced? Where did the energy go then?

Just to make an analogue. Imagine that you're stitched some reactive engine to some cart without wheels and trying to move it. Lots of friction, so lots of energy will convert to heat. That's basically insulator. Need a lot of energy to go through it.

Now you add wheels to the cart. Well, it moves quickly now. Though still you have some friction and some heat. That's conductor.

Next: you add wings and don't need wheels after lift-off. Only air friction. Still some energy converts to heat, but speed and efficiency much higher. That's copper, one of the best conductors.

Now remove air, you're in the space. Well, no friction, you can accelerate as much as you want. So this is superconductor in a nutshell.

There's nothing fundamental about resistance. Yes, most elements are not superconductors, but some are superconductors and they require not very exotic conditions. So there's nothing groundbreaking about inventing room-temperature superconductors, physics won't be rewritten because of that. But lots of practical applications, of course, so it's incredible useful invention if true.

I don't know how much CPU efficiency can we squeeze theoretically. There will be losses, even if everything is superconducting. Radio waves, for example. And it's not clear if it's possible to make everything superconducting.