Looking at the mess regarding GPU and ram, I start to wonder whether we’d be able to produce enough supra conductors and have enough helium to support quantum computers leaving research lab and being industrialised?
While I doubt quantum computer would be massively produced, I am not sure how much materials are needed for a quantum processor compared to let’s say, an MRI to take the most common supra conducting machine
I presume we’re talking about superconductors; I don’t know what a supra (?) conductor would be.
There are two questions here: 1) how much superconducting materials are required for today’s state-of-the-art quantum computers , and 2) how quantum computers would be commercialized. The first deals in material science and whether more-capable superconductors can be developed at scale, ideally for room-temperature and thus wouldn’t require liquid helium. Even a plentiful superconductor that merely requires merely liquid nitrogen would he a bit improvement.
But the second question is probably the limiting factor, because although quantum computers are billed as the next iteration of computing, the fact of the matter is that “classical” computers will still be able to do most workloads faster than quantum computers, today and well into the future.
The reality is that quantum computers excel at only a specific subset of computational tasks, which classically might require mass parallelism. For example, breaking encryption algorithms is one such task, but even applying Shoe’s Algorithm optimally, the speed-up is a square-root factor. That is to say, if a cryptographic algorithm would need 2^128 operations to brute-force on a classical computer, then an optimal quantum computer would only need 2^64 quantum operation. If quantum computers achieve the equivalent performance of today’s classical computers, then 2^64 is achievable, so that cryptographic algorithm is broken.
If. And it’s kinda easy to see how to avoid this problem: use “bigger” cryptographic algorithms. So what would quantum computers be commercialized for? Quite frankly, I have no idea: until such commonly-available quantum computers are available, and there is a workload which classical computers cannot reasonably do, then there won’t be a market for quantum computers.
If I had to guess, I imagine that graph theorists will like quantum computers, because graphs can increase in complexity really fast on classical machines, but is more tame on quantum computers. But the only commercial applications from that would be for social media (eg Facebook hires a lot of graph theorists) and surveillance (finding correlations in masses of data). Uh, those are not wide markets, although they would have deep pockets to pay for experimental quantum computers.
So uh, not much that would benefit the average person.
Modern cryostats don’t need a constant supply of helium because they work on a closed-cycle principle in the same way your fridge does.
Irrelevant, I think? Whether we have enough or not, they’re going to just take it. See: the “AI” boom and how SSD and memory costs are skyrocketing and general availability declining as well as GPU costs/availability since Buttcoin.
I’m not too worried now, but at some point, there’s going to be a breakthrough and “quantum” is going to be the new bubble. “Quantum datacenters” and/or “quantum AI” or whatever marketing horseshit they come up with.
