Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Physicists create elusive particles that remember their pasts (quantamagazine.org)
102 points by peter_d_sherman on Sept 12, 2023 | hide | past | favorite | 27 comments


So did they create particles or just simulate the creation of particles on a quantum computer?

> Some physicists consider these efforts simulations, because the qubits inside the processor are abstractions of particles (while their physical nature varies from lab to lab, you can visualize them as particles spinning around an axis). But the quantum nature of the qubits is real, so — simulations or not — the processors have become playgrounds for topological experiments.

So far (AFAIU) anything that can be simulated on a qc can also be simulated (much more easily) on a classical computer. Did the simulation leverage specific quantum properties of the qubits in a way that resulted in a genuinely never-seen-before physical system, or did they just simulate it on a qc for the heck of it?

Guess I'll wait for the Scott Aaronson writeup.

Edit: feeling kinda bad to give the stereotypical skeptical/dismissive comment, so I want to stress that I legitimately don't understand a lot of what is actually happening here. Maybe it's way cooler than I'm making it out to be.


This Quanta piece is from May and is talking about work that came out in October 2022 so if Aaronson was going to write a piece he probably already did (maybe, obviously my assumptions could be completely wrong).

But the abstract of the pre-print (https://arxiv.org/abs/2210.10255) covers some of what you are asking:

"Despite the well developed mathematical description of non-Abelian anyons and numerous theoretical proposals, the experimental observation of their exchange statistics has remained elusive for decades. Controllable many-body quantum states generated on quantum processors offer another path for exploring these fundamental phenomena. While efforts on conventional solid-state platforms typically involve Hamiltonian dynamics of quasi-particles, superconducting quantum processors allow for directly manipulating the many-body wavefunction via unitary gates."

They created a collection of quasi-particles that has different statistical properties that we don't see in 3D (the Non-abelian anyon https://en.wikipedia.org/wiki/Anyon). So simulated or created becomes a tricky word here, the quantum processor is putting these qubits into a state that acts as a quasi-particle so they can study it directly. So no a classical computer would not be able to do this in the same way it would have to use classical bits to simulate the quasi-particle.


Thanks for the explanation.

That quote is extremely terse and would have taken a considerable amount of time to understand.


I believe the blue figure in the middle of the article shows such a 'simulation'. It is called a simulation because quasi particles (ie groups of physical particles) are treated as qbits.


I haven't read the article but the point is that a quantum computer is not a simulation of a quantum system, it is completely a quantum system. If you can prove that the dynamics and states of your qubits can be mapped onto the system you want to simulate, then it is physically equivalent, not just a perfect simulation of it. In a sense, the word simulation is a bit too weak to describe what is actually going on here, possibly the phrase "more convenient example" would be more accurate


It's more like using an electrical circuit to make an analogic simulation of a spring with a mass (or a pendulum). You get the same simplified equation, but each one has different nasty corrections. In the case of an electric circuit, you are usually ignoring radiation, thermal noise, ... In the spring and mass you are usually ignoring the non linear part of the spring, air viscosity, tidal forces of the moon, ...


It's like using an electrical circuit to make an analog simulation of the behavior of an arbitrary resistor-capacitor circuit.

You can get a lot of flexibility from all the extra stuff you have controlling your simulator, but the simulator is essentially the same thing it is simulating. You don't have to ignore anything.


A quantum example of that is how the light in a cavity is exactly the same as the mass on a spring but with a temperature independent "thermal noise", but only if you consider just one mode of that spring, ignore non-linear stuff, etc.


interesting, some kind of induced physical analogue, or induced physical form or something mirroring, behavioral state mirroring, physical state mirroring, yes what do we call this phenomena. Has almost the feel of how wave energy can be found in all sorts of media, fluid, sand, gas, etc.


>So far (AFAIU) anything that can be simulated on a qc can also be simulated (much more easily) on a classical computer.

A few things about this idea.

1. It really depends on what you mean by "more easily." From a technical perspective, yes, simulating this on a classical computer is much easier. But simulating this on a quantum computer gives an exponential speed up, presuming you can live with the provisos associated with such a simulation.

2. Suppose our _model_ of quantum mechanics itself is wrong. For example, suppose we live in a universe where some peculiar physical collapse theory obtains (it would have to be peculiar indeed given the constraints we have on such a thing, but presume anyway, for the sake of argument). Then the quantum simulation may indeed tell us something that we don't know. If you've read Aaronson's book, he sort of suggests this is one of the cool things about Quantum Computing: it operates as an experimental domain within which some pretty strict limits on our quantum theory can be tested. Eg, if we did discover that there is a physical collapse of some kind, then one way that might happen is with a non-linearity in the Schrodinger Equation. But if there is such a non-linearity than it has some pretty profound (one might even say absurd) implications for what you can get away with with a quantum computer.


From what I understand most qubits are implemented as trapped ions or quantum dots which are both collections of particles instead of a single particle. For quantum dots you may have thousands or millions of atoms in a single dot, since they are nano crystals. Their behavior is quantum so they can be called “particles,” but not in the same way an electron or quark are fundamental particles.

The simulation is like a reverse of those scaled flood models. Yes it’s a scaled simulation, but the same fluid dynamics come into play. It’s not a math model simulation like you would perform on a classical computer.


Don't feel bad. It wouldn't be the first time Quanta engaged in such practices.

https://mateusaraujo.info/2022/12/01/the-death-of-quanta-mag...



what their saying is a simulation is real if it can happen...since you can make a simulation it is real


Sounds like a medieval argument for the existence of God.


Cool. Let us software guys know when we get computers that are 1000x faster so we can write the next language/framework that is 1000x slower.


I can't picture quantum computers. They make no sense to me. Logic gates are just baked into my brain.


Scott Aaronson made a comic with Zach Weinersmith that explains quantum computing in an entertaining way:

https://www.smbc-comics.com/comic/the-talk-3


Wow!


Once the map between entanglement and exponential parallelism is clear, quantum computing becomes a breeze /s. Seriously though, entanglement is the root of what makes quantum computing useful.

One subset of problems that are solved exponentially faster are those that intersect with tensor products. Entanglement essentially "is" a tensor product of the probabilistic state space of two qubits.


There is evidence of logic gate analogues in biological neural networks, so this might be an analogy or even a straight statement of a fact!


Do photons have "memory of their pasta" as well?

From _ and EM Wave Polarization Transductions (1999):

> Time As Energy and Why It Is Very Dense Energy: In addition to the three spatial polarizations of photons and EM waves, there is a very, very useful t-polarization along the time axis. In this polarization, the 3-spatial energy is not oscillating at all. Instead, the time or time- energy is oscillating. Time can be taken to be energy compressed by at least c2, so it has at least the same energy density as mass. In other words, one second is 9x10^16 joules of time- energy (energy compressed into time). The t- polarized photon or EM wave is called the scalar photon or scalar EM wave, respectively. [... MKS units ... 1999 ... probably real]

Circularly-polarized light: https://youtu.be/QCX62YJCmGk?si=9Vqx7RB6v74WG5bv

"Physicists use a 350-year-old theorem to reveal new properties of light waves" (2023) https://news.ycombinator.com/item?id=37226121


Note that they don't remember all their past (like a gps device with a memory). They just remember how many times one of them has looped around a second one. This is well defined because they live in a plane. More details in https://en.wikipedia.org/wiki/Anyon


Is there only one past? Is the number of times represented as a discrete value or a function with some distribution? Is the time component one way, or could it make sense to remember how many times it will loop in the future?


So homeopathy was true! /s


But then it is true about everything not just water. \s


Cool




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: