Qubits based on FinFET at above 4 kelvin and use holes as spin qubits | Hacker News Reader

Qubits based on FinFET at above 4 kelvin and use holes as spin qubits

unibas.ch

It's kind of amazing to contemplate that the lack of an electron can, itself, have spin.

If you are in the field, it has probably become a commonplace. But it seems like it ought to mean something deeper about our reality.

Only in the same way that you can think of a bubble in a bottle of water as its own entity. When we describe the motion of the bubble, what we really mean is the collective motion of all the water particles that cause there to be a lack of water particles in the bubble region, but it's just more convenient to talk about the bubble as its own thing.

We can talk about the collective state of all the electrons that leads to a hole, or we can just pretend the hole is a particle and assign equivalent properties to it. There's nothing very mysterious about it, it's a modeling trick.

Right, it's just using the same book-keeping trick for both spin and charge. This:

    Hole with counter-spin + Bulk of electrons, all with spin

is much easier than this:

    Hole with no spin  +  Bulk of electrons but the one under the hole doesn't have spin

> There's nothing very mysterious about it, it's a modeling trick.

I wouldn't say it's just a trick. They have remarkable properties.

Quasiparticles (such as holes) have simple, relatively long-lived, non-dissipative properties like fundamental particles, such as momentum that is conserved along a path until an interaction, interactions with other particles, and a simpler wavefunction than would be expected from the combination of constituent fundamental particles.

They emerge from large numbers of fundamental particles acting together in harmony.

Like a bubble in water, except water doesn't harmonise that much, so it doesn't make a bubble with momentum (or spin) against the flow of water keep moving the same way until it collides with another bubble.

The 3-body motion problem produces chaos. It's a wonder that a 1000-body collection of electrons and nuclei holds a quasiparticle together instead of quickly turning into a non-linear, thermodynamic mess, and that interactions with other particles couple with the entire collection at once, to maintain the integrity of the quasiparticle.

At some level it's like a bubble, but the harmony is decidedly a quantum effect.

If it behaves enough like an actual particle, it calls into question the degree to which the things we think are actual particles really are.

thats because concepts like "spin" arent actually real but only a mathematical model. like all models, its a set of rules that allows us to make predictions, nothing more. its not reality.

A model that fails to model reality is generally useless, and something one hopes to avoid encountering at work.

Spintronics really are a thing! MRAM already makes use the spin of an electron for data storage and are some of the first devices along the path to wider use of spintronics.

My summary:

Scientists & practitioners found a method of building a quantum computer using novel techniques linked to proven technology based on industry standards (FinFET architecture) to build a quantum computer which offers a potential for scaling up to very large numbers of qubits.

"Our approach of building on existing silicon technology puts us close to industry practice," says Kuhlmann.

In layman's terms this group found a way to make a spin qubit, the smallest unit of a quantum CPU using either a positive ion sink hole or single electron with a negative charge.

FinFET may operate a low temperatures near absolute zero (0 kelvin or -273.15 degrees Celsius) but this design requires special cooling of the bottlenecks in the architecture.

Spin qubits store quantum information in the two states spin-up (intrinsic angular momentum up) and spin-down (intrinsic angular momentum down).

The rapid scaling of Qubits is thrilling and terrifying, notably for the implications on public key cryptography.

There are already public key schemes that are resilient to quantum attacks and I would be surprised if we build a scalable reliable quantum computer before we switch to these (not yet well tested) schemes. Old recorded communications will be decrypted but historically that has not been a very big deal.

When they say "resilient to quantum attacks" they just mean there is no known algorithm that could solve them efficiently. But since they aren't popular and we don't have quantum computers, not many people are even trying.

You are right! Just for completeness, I would add that bodies like NIST and various universities and researchers are putting significant effort into this work. I am indeed not qualified enough to compare their efforts to what the industry has already put in studying current encryption schemes.

We have not yet achieved even a single logical qubit suitable for breaking cryptography so it is impossible to estimate a scaling rate...

This is the most promising path to large-scale quantum coprocessors/accelerator cards. FinFET allows stacking billions of these electron holes in a CPU.

Delicious word salad anyone?

I think this translates to "technique to build a quantum computer using traditional silicon, maybe"

For a person in the field the title conveys a lot: It is a way to make qubits with traditional cheap scalable manufacturing techniques and, surprisingly, the qubits can work at relatively high temperatures that do not require 400k$ cryostats.

Yes this (plus using quite common machines to build/make a quantum computer)

$400k seems to be around what the helium recovery and liquidation to support medical imaging costs, but presumably at a far higher power than a $400k .1Kelvin cryostat.

4K is "liquid helium" temperature, as used in medical imaging. It is *relatively* cheap. Sub-1K is achieved by an out-of-equilibrium mixture of Helium-4 and Helium-3 (a much rarer, very expensive isotope). This is why sub-1K is much more expensive than 4K cryostats, for a given cooling power.

Integrated circuit qbits and readouts operable at surprisingly "high" temperatures.

Reply on news.ycombinator.com