Quantum Computers

[SkurjMaster]

Almost everyone assumes that a fully functioning ideal quantum computer will eventually be produced. Not everyone is agreeing with that. Let's supposed for a minute that the naysayers are correct and that a practicable quantum computer is not possible. What then? What is that the signal of?

[Avatar]

Some other method of improving computing power?

--A

[SkurjMaster]

What about the consideration that we are swiftly approaching the limit of silicon fabrication technologies and, thus, there's only so much improvement that can be made?

[Lazy Luke]

What if miniaturisation were not important? Having a bulkier computer - 3 or 4 CPU's, with several GPU'S - might for example boost computing power.
Like those security systems that run several screens simultaneously; breaking down a single screen using several GPU's would surely give better screen resolution!
As for cost, recycling the tech mountains poisoning the earth would be a good place to start.

[Rigel]

What about the consideration that we are swiftly approaching the limit of silicon fabrication technologies and, thus, there's only so much improvement that can be made?

We're already pretty close to that. AMD has chips coming out in the 7nm range, and that's small enough that quantum juxtaposition plays a pretty big role in how the currents flow through the chip (I think Intel started to account for quantum superpositioning of electrons around the 20-25nm mark, but don't quote me on that because I'm too lazy to look it up).

The next step for processing, like with storage, will be to go 3-dimensional.

Currently, most circuits are laid out in a 2-dimensional plane, because that's the easiest way to manufacture them. We started layering multiple planes for storage a few years back because the logic for storage is so simple (you only need poke and peek for a given address), but haven't really moved there for processing yet.

I expect that with consumers used to purchasing electronics at their current prices we won't see the massive jumps in cost that the massive jumps in performance could warrant; instead, we'll see processors that are comparable to the current ones but in a fraction of the space, and consuming a fraction of the power. Remember that Nokia phone you had 15 years ago that you only had to charge once a month? Yeah, imagine that, with your current phone... Certainly, it would be worth a few hundred extra, right?