18 ms·

Can we not simulate even one atom of oxygen well enough to determine if it's stable or not. What if they used a big computer like the ones they keep building at

• foota 3y ago
  Iirc the simulation is more difficult the more components there are in an atom, and oxygen 28 has a lot of them.
• moelf 3y ago
  simulating atom is hard because the interaction between protons and neutrons are described by not only EM force but also Strong force, and in particular, Strong force at this low energy (as oppose to what happens at the LHC) is hard, partly because https://en.wikipedia.org/wiki/Asymptotic_freedom https://en.wikipedia.org/wiki/Asymptotic_freedom , basically our usual physics tricks of "perturbation approximation" doesn't work because the Strong force diverge at lower energy
• coder543 3y ago
  From the article: “Oxygen-28 might prompt physicists to revamp theories of how atomic nuclei are structured.” If the theories are incomplete or wrong, how could we accurately simulate things we don’t yet understand? It doesn’t matter how powerful the computer is.
  • rcme 3y ago
    Isn’t that the point of simulation? To gain understanding of something? I think the underlying point is that, if our understanding is so incomplete that we can’t simulate a single atom, how can we trust all of our other physical understanding?
    • ars 3y ago
      No, that's not the point of simulation. You gain no understand from simulation because what you put into it is what you get out of it. Simulation can help you verify what you already know, and point to areas to examine, but you can't actually directly learn anything from it. We trust our physical understanding by experimentation, not simulation. You can use simulations in areas that are fully understood to run calculation on new arrangements of those those things, without having to make the physical object. But it only works when you already understand the thing, you can't gain that original understanding from the simulation.
      • nomel 3y ago
        I think there’s an assumption that the “unit” of simulation is much smaller than the oxygen atom, allowing these “emergent” behaviors to arise.
        • Retric 3y ago
          We don’t actually know the laws of physics, we just have some decent approximations. So sure run a simulation at whatever level of granularity you want that doesn’t mean it’s correct.
    • ChuckMcM 3y ago
      Yes and no. A simulation can test your model, which you then compare with empirical results to validate both the model and the simulation of it. When those results do not match, you have to figure out if your simulation is wrong or the model is wrong. The model said that O28 would be doubly magic and thus stable, and yet when they managed to make it, it was exceptionally short lived. As a result the model they have is missing something. On the plus side, now that they have an empirical result, they can tweak the model such that it continues to accurately describe what it currently describes, and describes a short lived O28. Once they have those tweaks, they can find another experiment to see if their updated model accurately predicts what the experiment would produce. If it does, they gain more confidence in the model, if it does not, they go back to tweaking the model. This is the essential core of scientific research, for science to be believable it needs to predict things that will happen given conditions, and then experimentalists establish those conditions and look for confirmation of the prediction. It is the only way to know if what we think we know is in fact worth knowing!
    • at_a_remove 3y ago
      You have this backward. So imagine you have a simulation, and you get an answer out. Yay. How do you know it is correct? You don't. You must compare against reality. Reality always wins. This is not a "single atom," you might as well say "a single person." Each one of those protons is composed of two up quarks and one down quark. Each one of the neutrons is composed of two down quarks and one up quark. Each nucleon is therefore three quarks, held together by the exchange of virtual quarks. The nucleons themselves interact via a stepped-down approximation of that called the strong nuclear force. And you're not allowed to forget the electromagnetic force, either. And then there's self-interaction ... There's a lot going inside of a nucleus. Simulations are only useful for testing your models.
      • cwillu 3y ago
        Also, the “composed of two up quarks and one down quark” is a dramatic simplification, kinda sorta like saying that the valence electrons of an atom are the only electrons. https://i0.wp.com/profmattstrassler.com/wp-content/uploads/2013/04/nucleons31.png https://i0.wp.com/profmattstrassler.com/wp-content/uploads/2... “Fig. 3: A more realistic, though still imperfect, image of protons and neutrons as full of quarks, anti-quarks and gluons, moving around at high speed. More precisely, a proton consists of two up quarks and a down quark plus many gluons (g) plus many quark/anti-quark pairs (u, d, s stand for up, down and strange quarks; anti-quarks are marked with a bar.) The edge of a proton or neutron is not sharp. Ignore the color-coding for now; it will become clearer in future articles.” -- https://profmattstrassler.com/articles-and-posts/particle-physics-basics/the-structure-of-matter/protons-and-neutrons/ https://profmattstrassler.com/articles-and-posts/particle-ph...
        • at_a_remove 3y ago
          Sorry, I meant to type "virtual gluons" instead of "virtual quarks."
      • wheelerof4te 3y ago
        "Each one of those protons is composed of two up quarks and one down quark. Each one of the neutrons is composed of two down quarks and one up quark." And we know this, how? Using magic? Has anyone ever seen a quark? We could barely detect atoms, now we're detecting something even smaller?
    • jacquesm 3y ago
      Because just like statistics don't say anything about an individual person they are quite useful when applying them to larger populations and we usually deal with larger populations of atoms. Gas simulations work well because the noise more or less cancels out and then your macroscopic gas laws emerge and allow you to say useful things about how a gas will behave. Even if under the hood it is a completely stochastic process that gives rise to these laws. The laws themselves are simply our best description of observed reality, they are not laws that any particular atom needs to obey!
  • addaon 3y ago
    There’s multiple levels of “theory” here. We have reasonable confidence that an ab initio simulation of a O28 nucleus would match experiment, but such a simulation is outrageously hard. We simplify things by creating an abstraction of the strong nuclear force, the residual force of the strong force at the scale of nucleons; there’s plenty of room for improvement here. Then, we have a further abstraction of “magic numbers,” a rule-of-thumb level theory that reduces the calculations of the strong nuclear force to a lookup table. While this last step is a pretty good approximation when applied to the EM force and electron orbitals, it’s no surprise that it’s a mediocre-at-best approximation for nuclear structure. Even so, finding cases where it doesn’t apply is useful for developing a refined version of this third-level rule-of-thumb — and a more accurate, more grounded rule of thumb here would be useful for refining speculations about the possible island of stability, where ab initial simulation is even less practical.
  • eikenberry 3y ago
    All theories are incomplete and wrong, that is a core principle of science. When and how simulations might be useful in testing that theory are context dependent. In other words.. it depends.
• semi-extrinsic 3y ago
  The strong and weak nuclear forces are insanely hard to simulate directly. You might have a look at e.g. this paper for some fairly current modelling: https://journals.aps.org/prc/abstract/10.1103/PhysRevC.103.034305 https://journals.aps.org/prc/abstract/10.1103/PhysRevC.103.0... Just to be clear, simulating the atomic nucleus isotope stability (like here) is something entirely completely different than simulating the quantum mechanics of electrons in one or more atoms (like we do in DFT), or simulating molecules (like we do e.g. in molecular dynamics). The latter two are comparably much easier.
  • ftxbro 3y ago
    it's paywalled
    • awesome_dude 3y ago
      Um, so I took the papers title and list of authors "Unbound spectra of neutron-rich oxygen isotopes predicted by the Gamow shell model J. G. Li, N. Michel, W. Zuo, and F. R. Xu" Threw that into Google Scholar and the only hit had a link to the pdf of the paper https://link.aps.org/accepted/10.1103/PhysRevC.103.034305 https://link.aps.org/accepted/10.1103/PhysRevC.103.034305
      • AnimalMuppet 3y ago
        [flagged]
    • moelf 3y ago
      https://arxiv.org/pdf/2103.01478.pdf https://arxiv.org/pdf/2103.01478.pdf
• ars 3y ago
  No, we can not. We can't even simulate a hydrogen atom, which is far simpler. For that matter we can't simulate a single proton either. See: https://www.quantamagazine.org/inside-the-proton-the-most-complicated-thing-imaginable-20221019/ https://www.quantamagazine.org/inside-the-proton-the-most-co... (Unrelated but this is why I don't believe singularities exist in the universe - we don't know enough about quark degeneracy pressure to know if it's actually possible for a star to collapse - it's possible the quark pressure keeps the matter from compressing.)
  • AtlasBarfed 3y ago
    They're just quark stars with an event horizon?
    • ars 3y ago
      No event horizon, just incredibly powerful gravity.
      • AtlasBarfed 3y ago
        Singularity or not, strong enough gravity produces the event horizon, does it not? It's the boundary where escape velocity exceeds the speed of light?
  • addaon 3y ago
    > it's possible the quark pressure keeps the matter from compressing Why are you more comfortable with infinite pressure forces than infinite densities?
    • AnimalMuppet 3y ago
      It wouldn't have to be an infinite pressure force - just enough to keep a finitely-sized collection of quarks from collapsing further.
      • addaon 3y ago
• rich_sasha 3y ago
  I was once told simulating N quantum particles interacting with each other is exponentially hard in N. I wouldn't know myself. But if it's true, that's why you can do your hydrogen atom in Quantum 101, and why this is not merely O(28^2) harder (or do you need the electrons too?).
  • db48x 3y ago
    Exponential means 2^N, not N^2. Also, notice that every proton and neutron is really three quarks, so its at least 2^84 times harder than a hydrogen atom.
    • shwaj 3y ago
      I think they intended to say quadratic. For N classical objects interacting, each of them can interact with each of the N-1 others, hence O(N^2). The GP is saying that the quantum interactions don’t follow this rule, hence “not merely O(28^2)”, but in fact much bigger.
      • btilly 3y ago
        No, they meant exponential because that is what they heard. That is correct, it is exponential. They were just apparently confused about what exponential means.
        • shwaj 3y ago
          Not to make a big thing about it, but you’re wrong. Try read it again, slowly. Edit: hint: focus on “if that’s true” and “merely”.
          • db48x 3y ago
            I disagree too. He asks “If it is truly exponential, why is it not just 28^2× (aka 784×) harder than hydrogen?”. The answer is that he misunderstood exponential growth, and thus it is at least 19342813113834066795298816× harder than hydrogen.
• tsimionescu 3y ago
  This is one of the tasks for which a functional quantum computer is needed. Simulating quantum systems with classical computers is (as far as it is known today) exponentially hard. Even simulating a hydrogen nucleus (which has a single proton made out of 3 quarks) is actually too complex for even the biggest classical computer we can build. An oxygen atom is FAR beyond what we could ever simulate with any known algorithm. Note: it's not currently proven that it's impossible for classical algorithms to simulate quantum systems in polynomial time, but it is strongly believed to be the case.
  • semi-extrinsic 3y ago
    Do we know any quantum algorithm by which a quantum computer would be able to simulate nuclear physics better than a classical one? The interactions inside a nucleus are completely different from regular quantum mechanics with electrons etc. like in a quantum computer.
    • btilly 3y ago
      https://www.pnas.org/doi/10.1073/pnas.0808245105 https://www.pnas.org/doi/10.1073/pnas.0808245105 shows them winning on multi-body atom problems. https://www.nature.com/articles/s41598-023-39263-7 https://www.nature.com/articles/s41598-023-39263-7 shows them winning on simulating a nucleus.
      • bsder 3y ago
        IF you can build a scalable quantum computer, it can do this. Note that "if"--it's a big, unsolved problem right now. The problem right now is that once you start adding qubits the noise in the system grows faster than your signal.
        • btilly 3y ago
          The question is whether the algorithm was known, and not whether we have the technology to actually implement the algorithm.
• Angostura 3y ago
  We did simulate it - using our models. The model said 'stable'. Reality said 'nope'
  • ftxbro 3y ago
    The simple shell model said 'stable' but we can calculate with more detailed models than that one.
• jacquesm 3y ago
  To help give a way to grip why this is a hard problem: computers work with quantized values so you get quantization noise (or you get thermal noise in the analog domain), and that means that every datum that you want to ingest, compute and spit out again as a result is going to be about right, but not quite right. You then take it through a few million cycles to see how it behaves and as you do so with every iteration you are further and further away from how it really would behave. And that's just that one aspect: quantization noise. Then there is the uncertainty principle to deal with which may preclude one or more parameters from being known exactly in the first place. And so on. In the end you find that no matter how much computing power you throw at it even just a simplest atom is beyond your capability of simulation for as much as a tiny fraction of a second. What we do in almost every simulation is to take a shortcut: instead of simulating the individuals atoms we simulate their observed properties and usually in larger numbers. This allows for useful work to be done in a timespan not measured in aeons. But it's an approximation at best, never a simulation accurate enough to make definitive statements about how any individual atom behaves and what its future state will be given some set of initial conditions with any accuracy.
• bazzargh 3y ago
  Speaking as someone whose PhD was simulating the nuclear structure of nuclei like this... yes, we probably could, now. Back then (like, 30 years ago, and I stopped doing physics after this so my memory of this is fuzzy) we were looking at simulations of nuclei like O16. I say 'we' - nuclear theorists were _very_ thin on the ground, we were the only remaining group in the UK. Most particle physicists are of the kind looking at subatomic particles, not nuclei. Anyway, we were attempting to port the code to run on parallel processors (a 96 transputer rack at the time), and then diagonalise the matrices of the interactions to get out a spectrum of energy levels. IIRC the matrices worked out as ~20m x 20m, and the technique used was the https://en.wikipedia.org/wiki/Lanczos_algorithm https://en.wikipedia.org/wiki/Lanczos_algorithm ... the problem we had was that the state space explodes combinatorically with increasing numbers of nucleons; and the computation time scaled something like n^1.1 for n states, due to inter-processor communication. In the end that was what killed the project - it became clear that with moores law we were about 10 years from having affordable access to a computer that could do the calculation for larger shells (including O28, which was well out of our range). That was the state for _exact_ calculations, but there were alternative approaches - I recall VAMPYR being a German Monte-Carlo simulator for shell models that performed really well, and could extract properties even for quite heavy nuclei. Looking back a lot of the problems were just a lack of memory, even more so than compute. The matrix elements weren't stored explicitly but recalculated on the fly because we lacked memory, this led to us not using off-the-shelf matrix code and the whole thing had numerical stability issues and used lanczos because we could fit that into the memory on board the processors. These days I use servers in AWS with ungodly amounts of memory and extremely fast cpus, I'm pretty sure they could simulate this for a couple of hundred bucks.
  • sjtgraham 3y ago
    This is the kind of HN comment I live for.
  • jacquesm 3y ago
    Can you give an indication of the amount of precision required for the initial conditions to be able to perform such a simulation? Fascinating stuff this, thank you for your comment.
• evanb 3y ago
  16Oxygen is well-characterized using a variety of ab initio nuclear many-body strategies, partly because it is doubly magic and has an equal number of protons and neutrons. Even though 28O is doubly magic, the unequal number of protons and neutrons creates computational difficulties for even ground-state computations, never mind decays. Doubly magic: https://en.wikipedia.org/wiki/Magic_number_(physics) https://en.wikipedia.org/wiki/Magic_number_(physics)
• pfdietz 3y ago
  No, we cannot simulate nuclei that accurately.
• akomtu 3y ago
  Contents of the nucleus is the same level of mystery as contents of a black hole: we know the basic bulk properties, we know what happens when it collides with various particles, but what's going on inside is a matter of speculation.