JackGriffin wrote:
One of the defining problems for the processor chip is the speed of light (of all things).
I have to correct something here because that's a huge misconception of things, because I think you're not seeing how fast light travels, it's 300.000 Km per SECOND (or 186.000miles/s, slower of course in circuitry, but just slightly). Light is not "that" slow at all and is NOT the reason why a processor is faster or slower, 200 years or more will still pass until we reach the point that we get limited by that.
What limits above all the speeds within a processor or any chip for that effect, are the semiconductors themselves, since they have a response time.
For instance, perfect square clock signals do not exist, as a perfect clock signal would go instantly up and then down forming a perfect square in the sign change, perfect 1 and 0 if you prefer, but in reality they take some time to go up and then down at those moments, and that is the response time of a semiconductor component, and that is what limits any signal processing nowadays (there are other factors, like interference and the way you turn lines within a circuit because of electric fields, but this one is the biggest).
If the speed of light was limiting this, I sure you that running 10 or 20 or more instances of UT3 in one of those processors would take less time to process than an "Hello world" output in a today's i7 using machine code. In a processor where the light speed is a limit, running any heavy program nowadays would be a joke to that kind of processor.
By shortening the distances, they get less loss in the signal, a better response time, less unwanted electric fields, because there's less material that influences this timing. Plus, at those sizes the electric current is minimal, precise and prone to losses, so less distance also means less power which leads to less heat and more stability, because any nm of connection makes a difference at extremelly low current usage.
Another big misconception is about electricity itself: when you turn up the lights you just turn a stopped highway of electrons to move again, because electrons aren't free, they are actually fighting for a spot in valence electron locations for an atom, and since you create a diferencial of voltage, and the the whole system has to balance again, so it's all bound to the response time of the material itself in a closed circuitry. What many people think, and wrongly, is the "signal" has to reach some point first, but this is not a pipe system and water flow, nor light itself, electricity works differently.
An electron may move at the speed of the light, but it doesn't need to reach the other side so you can get an electric signal, you just need electrons to move.
Sorry for the big post