Quantum computing is coming close to reality, and its applications seem almost boundless. However, the bounds of those applications and of the very atoms involved are real, with research bridging quantum energy, uncertainty, and newly discovered limits of just how fast a quantum computer can get.
Dr Gal Ness discusses his research into quantum speed limits opens up new understandings of the cutting edge of physics.
Read the original research: https://doi.org/10.1126/sciadv.abj9119
Image Source: cybermagician / Shutterstock
Transcript:
The following transcript is automatically generated.
Will Mountford:
Hello I’m Will. Welcome to ResearchPod. Quantum computing is quickly progressing from an idea out of science fiction to an ever-nearer technological reality, and its applications seem almost boundless. However, the bounds of those applications and of the very atoms involved are real, with research bridging quantum energy, uncertainty and newly discovered limits of just how fast a quantum computer could get. Today I’m speaking with Dr Gal Ness about how his research into quantum speed limits opens up new understandings of the cutting edge of physics. Dr Ness, hello. Hey, Will, Thank you so much for your time and joining with us today for a little bit of reference for everyone at home, for me as well, Could you tell us a bit about yourself, your work, your research and some of your academic path that’s led to where you are today?
Dr Gal Ness:
Yeah, so I’m from the UfSegist Group at the Technion Haifa Israel. The work that we’re about to describe today. This is one of the subjects on which I focused on my PhD studies, which is quantum speed limits, and the experimental setup in the University of Lisbon offered us a unique platform to really push the limit on how far, experimentally, can we study.
Will Mountford:
Now I’ve got to say the phrase pushing the limits on experimental studies of quantum speed limits is one of the most sci-fi kind of sentences I’ve heard in a very long time And I’ve got to know how does it feel personally to be working at the very cutting edge of how we understand how the universe works?
Dr Gal Ness:
Well, always when you’re zooming into the world, you feel like I have this very little riddle that I want to study. But of course it fills you up with very great enthusiasm when you zoom out a bit and you realize that your work or your day job is basically measuring something, usually something that has never been measured, and then trying to baffle with it And it doesn’t make sense. And you try to take it to collaborators and working out a bit from all directions.
Will Mountford:
Zooming out from that riddle, to put it from the micro scale to the macro scale. How are you kind of connecting those tiny, tiny measurements with the rest of the world around you, perhaps?
Dr Gal Ness:
Yeah, so the pleasure of working with the systems that we work with, which are basically ultra cold atoms, is that they are maybe very complicated experimental apparatus But in some sense they are very, very, very simple. So we take atoms, which are the fundamental elements of this universe, and we play with them really, really gently, and this actually gives us the ability to do simple manipulations and study general properties of you know objects around us.
Will Mountford:
They mentioned some of the international collaborators you’ve had on this project And I’ve seen, looking over your profile, that you’ve been presenting at conferences all around the world and winning awards here, there and everywhere. How would you describe the international community working in quantum speed limits and quantum computing?
Dr Gal Ness:
So basically, it’s really a great place to be in academic terms just a wide variety of fascinating people to learn from and share with your findings, but mostly even this specific concept of quantum speed limit. Even within this community it’s very fundamental on one hand And on the other hand it’s a previous experiment of research for quite narrow. So I feel quite lucky to present this kind of concept within physics conference and they received quite a bit of attention And you know it was really nice to find interest for other physicists learn about this issue.
Will Mountford:
Okay, and to be a window for the audience and for myself as we get into the details of your work, maybe we can start by, you know, opening up the scope of the interview and setting the scene with why, out of everything in the world, out of any other topic in physics, out of any other topic in science, why quantum computing. what is it that matters about this work to you and how you think that matters to the world at large?
Dr Gal Ness:
Well, the first motivation I would say even not mine, the founders of direction is it’s quantum computing on a very fundamental level is simply hard. It’s really, really difficult to try to simulate a quantum phenomenon on a classical. In fact, we are very limited in our abilities to. So nice idea you to reach out to Feynman was. Well, maybe don’t try to hammer this with a lot of computation, a lot of classical computation, because it’s really intractable. But let’s try to build some machine which could be a computer that is based on these fundamental principles of quantum mechanics, and this is of course, not an easy task to make. But the prospect of building such a machine I really endlessly can really lead us to learn a lot about our world and also on the computational sense. The feature great advantages over classical, something which really drives to. You know, hit this.
Will Mountford:
Well, let’s get into some of the details of your work. You probably Lay a good foundation in terms of some of the language that we’re going to be using to discuss quantum computing and some of the terms that go with it. So in observing and measuring quantum phenomena, like you say, there is the whole measure of uncertainty, which some people might be familiar with as an idea. We also gonna be talking about qubits, gates and energies. So if you just get a glossary of terms to start with and then we can use that to explain the work that you’ve been doing, Sure.
Dr Gal Ness:
So let’s start with classical objects, with what we’re more familiar with, and try to grasp the concept of uncertainty. So in classical systems uncertainties usually a measure of class of object. Take, for example, this pencil. And if I take one pencil out of my drawer and measure it about 10 centimeters long and I pick another one and it’s just a bit more and the third one will be a bit less. So I can say that the average length of a pencil it’s about 10 centimeters and I have some error or some uncertainty of few millimeters. So if I don’t know A priori you’re at the length of a pencil, i have some uncertainty of its length. So in the classical order is the property of distribution, of cluster of objects. In the quantum world we can have this distribution within a single object. So let’s say that we have a quantum pencil Inherently feature several links which might be 10 centimeters, 11 centimeters, and once we perform something we call a measure or a collapse of its way function, then we set on a specific link. But this distribution, this uncertainty of length of any other measure, is something which is inherent to the quantum object.
Will Mountford:
And then when it comes to cubits, gates and energy. So is a lovely blog that came up through some of the other work about marbles in a bowl of light that kind of illustrated them in a very poetic and visual way. But if you explain it for our audience listening to this now, what some of those terms mean as well, Yeah, so it’s hard to be the cubit.
Dr Gal Ness:
Cubit is basically the fundamental building of a block of quantum computers. Cubit is just the quantum equivalent of classical Be it. So in quantum we have cubit. Beats means that you have two values. You have either zero or one, and single, for cubit has either zero, one, let’s call it red and blue. But what’s special about cubits is they don’t only have this zero and one, you can also have like 5050, so the particle can split between the two or you know any other population of the two. But it’s not like As we said in the earlier example, it’s not a property of distribution of particles, but it’s actually the very single Particle that has superposition or being at the same time at the two states. So between the zero and the one, we’re not talking about a measure of 0.5, but zero and one at the same time exactly, and if you will repeat the measurement many times, then you will find that some of the times it will be one and some of the time it will be zero.
Will Mountford:
Okay, and then the gates. Part of that.
Dr Gal Ness:
Yeah, so so now you have this cubit. You said I like beats, but have more fancy properties. You want to pass them, just like you know. You want to harness these features and perform computation. So how do you do computation? you say, okay, i have your few beats. This is zero, zero one. If the fourth one is zero, make it one. If it’s one, keep it, and so on and so forth. So quantum gates basically Perform the same act, just that they can even play with the inner probabilities. So we can say, if we have an atom which is 80% at one and 20% zero, flip it around or some other manipulations that you can do, and basically this gets allows to perform computation. Now, when we perform a sequence of gates, this is basically computation, right, also in the classical sense. And this sequence of gates can be characterized by something we call a distance in the quantum. So if I want to ask what is the speed of quantum evolution, i can think of some computation which I will define as a distance, and ask well, how long is this? And also, how long does it took to execute all these gates? So, what’s the time? And, as we remember from elementary school, you take distance divided by time and you get velocity or speed, and this is basically the speed limit, the quantum speed, the amount of computation you do divided by the time.
Will Mountford:
In my head. I’m picturing this as like a series of I watch on YouTube the marble run Olympics, where a guy is like pinging marbles down, like long runs of tubing to like sit, basically come up with an Olympic table of how fast his marbles are moving. It’s ridiculous, It’s absurd. I love it, but this is how it’s kind of visualizing in my brain. It’s not that full And lastly, the idea of interference.
Dr Gal Ness:
So interference is maybe the most fascinating phenomenon about these qubits or quantum, And this means that we have these different distributions or population of states within superposition, So which are to meet particles. Each object have like non deterministic state And we can perform manipulations, for example gates that actually share information or make two of these objects interact. So interference is the act when the populations of probabilities of one quantum and affecting the population of another one, And together we can extract some useful information about the relation.
Will Mountford:
And with all of that said, do we feel like we have enough covered to start addressing your particular experiments?
Dr Gal Ness:
Maybe just to relate to our previous example of the length of the pencils. I will just emphasize here that what we are interested in is actually energies of levels. So if we have zero and one, for example, it’s a cast to two energies, so two values of energy that the atom can be in, or particle or whatever, and it can even feature more level than only two. So going beyond the regime, that’s basically what we did in the experiment, why the results were so interesting And the reason we are interested in energies and in the uncertainty of energy and the mean value of this energy is that these values would dictate the quantum speed limit, as we will see.
Will Mountford:
So, with the establishment that there are speed limits, can you tell us what part those speed limits play in the setup of your experimental observations, the lasers, traps and cooled atoms, part of things like the physical operation, how you are making these observations and doing these measurements.
Dr Gal Ness:
So we are using cold atoms which enable us to study very carefully this evolution of the quantum state, and these atoms are actually extremely cold. Just to give some kind of feeling to how cold they are, so worth mentioning that there is an absolute zero temperature in this world, where everything kind of freezes about minus 300 degrees Celsius, and in this scale you can talk about how much we are hotter and colder than other objects. So, for example, we are 20 times colder than the surface of the Sun, more or less. And if we can think about liquid nitrogen, for example, something like widely known but still very cold, then we are something like only five times warmer than this liquid nitrogen, while our atoms are something like 50 million times cooler. So they are really really extremely cold. And the reason we go over all this effort and make them so is that in this regime of temperature, instead of feeling around like the atoms in the earth around us in jet plane speeds, they almost stand still. So we can really delicately study them, so we can get rid of all this thermodynamic motion that we are not interested in. So these atoms enable us to study very carefully the bounds and we can also create very precise potentials that hold these atoms. By these potentials, which we can analyze, their energy spectrum or their energy landscape, we can execute measurements to probe these quantum systems.
Will Mountford:
Now the probe, those measurements that you mentioned. I believe the term that I’ve picked up from your papers is a fast excitation interrogation scheme. Is that right?
Dr Gal Ness:
Yeah. So I think I will not dwell too much on the technical term, but what’s the essence of the experiment that we did is that you want to study the dynamical properties of these quantum systems, and this is non-trivial. Usually, when we think about quantum entities that we want to measure, we arrive at some final state and then we want to characterize And here we try to characterize some motion. So when you try to capture dynamics, you need your probe, you need your measurement tool to be much faster than these dynamics or much finer at the temporal scale. So we made an effort to execute these gates, these interferences that we induce on the atom, on a very fast time scale. The time scale of evolution that the atoms undergo in these experiments are something on the order of microsecond, which is basically one millionth of a second, and what we did is to build on Raman transition, which is a very short elaporado to perform the executed these gates in something less than 50 nanoseconds, one billionth. So to give some sense of these Raman transitions so vital to this experiment. So the usual way to translate atoms between two states, the two states that we are interested in here is using microwave antenna, just like the microwave antenna used to warm up your lunch, but if you want to drive transitions fast with microwave, you would probably need to invest so much energy with this direct transition that you will cook up the whole issue. So if you would like to avoid this, you could use this Raman transition to utilize light, and this light is something that we can focus directly on the atoms, and by this we can induce very high intensity or very fast gate rate directly at the position that we are interested in and implement this fast transition. I just want to say that basically, these Raman transitions are the physical reason for me to move to the University of Bonn for active collaboration of few months there in order to actually construct the optical system to perform this kind of Raman transition, and in this context, i would really like to acknowledge my partners, especially Manolo Rivera-Lam, who worked with the optical setup, and the other teammates which made this.
Will Mountford:
We said quantum state evolution a couple of times there, and that is the state of approaching certainty within the bounds of between zero and one, or the movement from one state to another, With the scale of temperatures that you’ve set there somewhere between the sun and absolute zero. it would be nice to think that you could use those cooled atoms for a functional purpose, And that’s what your experiment relies on, is having them and using them to make those measurements right.
Dr Gal Ness:
Yeah, so basically we take these cold atoms and the reason we execute this experiment, the specific laboratory at the University of Bonn, what the specific feature that their experimental setup is offered, is something which is called a spin-dependent optical lattice. By that we mean that we have some potential which we can imagine as some hill or some slope that the atom can roll along. And this potential is actually different to different states within the atoms. So if we imagine our two states as red and blue, we can think of the atom in split into two marbles, these two colors, and we have a red slope for the red marble and the blue slope for the blue marble. And now, because we can position them separately, what we can do is we can prepare the atom at a position which is at the bottom of the blue slope, so the blue part of it would stay still while it is on the top or somewhere in the middle of the red slope. So the red marble or the red part of the atom would roll down the slope and evolve in time. That’s what we mean by evolution. So here we try to capture it, imagine it only as a physical evolution, like a mechanical one, but it captures most of the essence of also the quantum world, which is just slightly more complex. And then when we say evolution, we kind of ask how far does the red marble went? and if we want to ask how far did it went, we need to somehow compare it to where it was. But because these are quantum states, we cannot measure them twice right, because these are very fragile states that will collapse if we do So. That’s why we invoke this blue marble. To begin with, the blue part of the atom would stay still and serve as a reference for the initial state. So when we compare the two parts by executing another interfering gate, we can actually learn something about the evolution of the evolving part throughout the system.
Will Mountford:
Your work builds on a couple of decades of prior research into quantum mechanics, quantum computing and quantum state evolution. So to kind of backtrack a bit and kind of fill out the theory of what has led to your work on quantum speed limits and the experimental setup that you described, there’s some names and some Nobel prizes that would be good to appraise the audience of. The previous limits that have been set on quantum operations include those of the Mandelsten, tam and Margolis Leviton bounds. So could we cover what those bounds are, those limitations on quantum observations, and how your work on speed limits kind of fills out the next stage, the next part, the unknown area of what has been previously researched?
Dr Gal Ness:
So the basic question of quantum speed limits is well, how fast can a quantum state evolve? And the first clue that we have towards answering this question is something called the Heisenberg Answered in Principles. Heisenberg Answered in Principles is almost a mathematical principle, although it’s a very fundamental element of quantum mechanics which relates uncertainties of different measures of conjugate variables. So, for example, if you measure the position and momentum of a given particle simultaneously, your uncertainties are ought to be greater than some constant number related to the Planck constant. And this can also be translated into energy and time. But this yet not provide you with a solid or very refined limit on how fast the evolution can occur. Just give you the correct timescale. And what Manlstrom and Tamm did back in 1945 is that they formulated the relation between the energy uncertainty and the rate of evolution. They said, ok, given this and such uncertainty in your energy, the fastest way a quantum state can go from point A to point B is given by some formula. Later another bound was found by Margulos and Leviti which related this minimal evolution time to the value of the energy itself, to the average value of the energy. So we see that we have these two energetic values, this average value and the energy uncertainty. And we have these two limits, due to Manlstrom and Tamm and due to Margulos and Leviti, that constrain them and relate them to the fastest speed that can be under too quick quantum process. And these are, of course, two theoretical results. What we offered is an experimental observation of these two bounds.
Will Mountford:
And a third bound that will be known as the nest limit from now on.
Dr Gal Ness:
I highly doubt that, before we go to the third bound, that there is one thing else that we should mention. An interesting aspect of these two limits is that actually they have different functional form. What do we mean by that? Importantly, we said that we have these two bounds, one related to the mean energy and one related to the energy uncertainty. And now you can ask me OK, but what is the speed limit? So the speed limit is given by the most restrictive of the two, right. Every time one would be more restricted and it would provide the bounds. But interestingly, these two limits don’t evolve in time at the same functionality. Namely, if you have, for example, energy which is greater than your energy uncertainty, the Manlstrom-Tamm bound would always be the dominant one And your evolution would be governed by the Manlstrom-Tamm bound. On the other hand, if the energy is lower than the energy uncertainty, then we would see two phases of the evolution. First, the quantum evolution would be restricted by the Manlstrom-Tamm bound, as it was in the other case, but then only at non-initial times, it would translate to be restricted by the Margolis-Levitin bound. So you have two distinctive dynamical regimes to explore.
Will Mountford:
And how does the new bound that you’ve discovered fit into that picture?
Dr Gal Ness:
These are the dynamical regimes that we were probing in our experiment, and we found out that there is something very asymmetrical in this picture. So for most practical applications, where we actually make the system, the reason, inherent, upper bound on the energies. By this I mean that you cannot have any value for the energy. Although we have distribution, because it’s quantum and it’s unknown and everything There is still some cut where you say, okay, i cannot have energy greater than this and that value. And if you consider this energy, you find that even if you just randomize an arbitrary state, it’s almost all the time fall into the Manlstrom-Tamm regime, the single time scale, and this was not clear why this is the case. Then we thought, okay, but if you flip the world around, so you make this ceiling the top energy being your floor, we just invert all the energies. You find out something really nice. You find that there is something which is like the Margolis-Levitin bound that was before to the very small energies that were near the floor of your energies, but now what we now call floor is the ceiling. So basically we found another bound which is like the Margolis-Levitin one for these high-lying energy values. So for states that have energy which is close to the top available energy, there is a third bound which is very much like the Margolis-Levitin bound which we have for the low-lying and near the ground state. So in some sense, by analyzing this picture and observing this asymmetry between the two bounds, we discovered that there is a third bound to be addressed and this bound we call it the dual Margolis-Levitin bound because it’s kind of dual to it in its form. And using these three bounds we have a unified picture of quantum speed limits, which is very useful because what it gives you is actually by utilizing very modest knowledge of the state, so you only need to know the energy and the energy uncertainty and maybe the energy from the ceiling You get some bounds on the rate of evolution, which is something information-wise is very, very expensive. So in some sense, these quantum speed limits and the exploration of quantum speed limits helps you a lot when you try to draw estimations about complex quantum calculations.
Will Mountford:
With that new bound, that new limited mind. Is it too soon, either in this interview or in your research, to start connecting that back to quantum information applications like data storage and the process of computation?
Dr Gal Ness:
So I would say that quantum speed limits are very readily connected to actual applications computations. The little caveat about it I don’t know if it’s right for is that the quantum computers that we have now we are able to fully simulate them. So why draw bounds for something which you actually know what its velocity is? But I hope that in the near future we would have these quantum computers which we cannot simulate classically, and then the quantum speed limits can become really handy.
Will Mountford:
So you’re ready for the next generation of that processing.
Dr Gal Ness:
Yeah.
Will Mountford:
The future, eventual, maybe not next generation, but next, next, next, next. with this information in mind that’s already being used, are there any other old experimental data that is changed or challenged by knowing that there is this third bound? is anything that needs to be revisited? prior research.
Dr Gal Ness:
So yeah, in some sense our previous experiment was only executed having the two first bound, that no one has revealed them at the point of the experiment. So, given this new bound and this new regime that measures three bounds, now there’s room for additional new experimental evidence can demonstrate the links with similar techniques.
Will Mountford:
so by inventing new techniques, Do you have any next steps, next research, new avenues in mind for how you would explore the adjacent space around this?
Dr Gal Ness:
Yeah, so basically related to this search bounds, you can more or less execute the same experiment and just study this new one. But more generally about one of the limits, i would say that a very interesting problem or very interesting direction to explore is something called open quantum system. So in the description to your end today, we focused on a quantum system that we completely know how to describe. So we have this, this and that distribution is. This solution is fixed Over energy until you measure it. But in reality we almost get to this case without very cold atoms. But in any other platform you always have some interference from the environment or some distorting interference. Another one that you like, and this interference can lead to very different dynamical evolution of the system. So there are already some theoretical attempts To construct, to formulate bounds for this kind of open quantum systems, and I would say that maybe the most interesting thing to do in experiment now is to experimentally probe This speed limits in open systems and see how this openness or relation to the environment of the system can relate To the question to cast a wider view, a longer view of maybe the next five, ten years, the next few, maybe generations of computing.
Will Mountford:
Where do you see this research going, fitting into the larger research environment, the international approach to quantum computing research? There’s any step beyond the next step that you think is going to be most interesting, or any end goal years and years down the line that you would like to see this be put towards reaching? You know you can discard the constraints of reality. No star trek teleporters or something.
Dr Gal Ness:
In some sense you could ask how applicable are these bounds, which is almost an equivalent question to how near future we would have quantum computers which are classically interactive. But, as I mean, in some sense, apart from this open quantum system, the quantum speed limits that we already explored are kind of the end of the story in this direction. So of course I am not a profit, but I don’t see how very specific research of quantum speed limits progress a lot without companion, which is which are full fledged quantum computer, because now it’s like we pull the theoretical idea, the ocean, and now it’s ready to be used. It’s waiting on the show to be used. It’s waiting on the show in some sense.
Will Mountford:
Now this is taken a lot of very sophisticated equipments and very specific setup to conduct these experiments. is it limited to just the investigation that you’ve done in the speed limit research or other places that you could go with us?
Dr Gal Ness:
We’re physicists so what we do is we play with our setup. That’s what we do for living. And we ran this measurements and found out results about one of the speed limits. But it was very intriguing for us to just let the system play for a bit longer. So not only ask like, okay, how far can this red marble get from the blue one? but maybe in our case it has some slope also, on the other hand, it and it will come back and the evolution might be more complex, so it might have some Revival or some periodicity in it. And we took quite long traces of this whole evolution of the particle for longer periods And this don’t have a lot to do with quantum speed limits, but they still very interesting measurement. And then we asked okay, we have this very precise measurement of the evolution, of the way function of this item. What can we do with it? and we ran and analyze this measurement And in fact we found out that by analyzing, by fitting these long time traces, of the overlap between the two colored marbles, we can extract a lot of information about the underlying state. Specifically, we can extract both the energy spectrum, so the underlying energy levels of the items, and the population in each level of the specific state. So this is something which is very unique. It’s kind of a graphic technique that we developed, and the reason we are so proud of it is actually it’s very model agnostic, so you almost don’t assume anything about the underlying state. You just give this long input and by fitting this long input to some very simple functional form, you can extract both the energy than the population, which is something very useful when we try to do quantum state analysis.
Will Mountford:
If people want to know more, where can they find your paper? and if they want to follow your research across quantum computing and anywhere else that your research leads you, where can they find more from you personally?
Dr Gal Ness:
Sure, so I have. I hold my personal publication record, a personal website, galnesgithubio, and there you can find basically all I published and also links to short talks about different papers. Might be interesting to you And if you are interested in this kind of subject, i very much encourage you to click on the very first link that takes you to your of the group website, my PT supervisor, which perform a lot of fascinating experiments in this direction of science.
Will Mountford:
For anyone who does want to follow up on your website, the TLDR summaries that you have all of your papers are tremendously informative and simplify things in a very accessible way. Thank you so much for those. Dr Ness, thank you so much for your time today and look forward to speaking to you again sometime soon.
Dr Gal Ness:
Thanks a lot
Leave a Reply