Transcript | Build Your Own Quantum Hardware Platform— with OpenQuantum

We’ve done plenty of episodes where there is a piece of software or cloud platform users can try, usually for the purposes of making coding or algorithmic development easier. Coding naturally seems to be the quickest way to get involved in quantum computing, but we’ve never discussed actually building quantum hardware at home! Find out how to start learning the engineering side of quantum information science by building your own neutral atom platform for about $8,000 or less! Join Host Konstantinos Karagiannis for a chat on what could be a new DIY path with Max Shirokawa from OpenQuantum.

Guest: Max Shirokawa from OpenQuantum

Konstantinos Karagiannis:

Over the years, we’ve done plenty of episodes where there’s a piece of software or a cloud platform you can try, but we’ve never discussed actually building quantum hardware at home. Find out how you can get started on the engineering side of QIS in this episode of The Post-Quantum World. I’m your host, Konstantinos Karagiannis. I lead Quantum Computing Services at Protiviti, where we’re helping companies prepare for the benefits and threats of this exploding field. I hope you’ll join each episode as we explore the technology and business impacts of this post-quantum era. Our guest today is the founder of the OpenQuantum project, Max Shirokawa. Welcome to the show.


Max Shirokawa:

Thank you. Thank you for having me.


Konstantinos Karagiannis:

We both spoke at Quantum Village this year at DEF CON. It’s a pretty hands-on conference for hackers. When we cover a do-it-yourself topic on this podcast, it’s usually a programming tool listeners can experiment with or something like that. But your talk stood out to me for reasons we’ll get to. Before we get to that, please tell our listeners how you found your way to the world of quantum computing.


Max Shirokawa:

My background is in software. Actually, I started in information security before I even went to college. I worked on the Red Team at Salesforce for a couple of years, and for a long time, I thought I was just going to stay in software.

I found it interesting. There were a lot of problems I like solving, but I was doing a part-time master’s at one point. I got a little fellowship to do that at Columbia, and I volunteered at this physics lab. They wanted somebody to write the control codes, and they were setting up this new ultracold-atom quantum simulator experiment. This is with Professor Sebastian Will, and it was very early-stage. It was maybe the third or fourth semester after he had joined the university.

We were starting to build everything from scratch. The budget, it wasn’t small, but it was lean, let’s say. As part of that, I had the opportunity to learn how a lot of these things were fabricated from the rudiments. At that point, I had not learned as much atomic physics or quantum mechanics or quantum computing, even, as I would have liked to. I was intimidated.

But what I found was that during the course of building the equipment that enabled the experiment without understanding the experiment itself, those mechanics were pretty accessible. That made the whole process less intimidating to me, because even if I couldn’t understand the theory that motivated the experiment, I could understand the mechanics of the tools you need to make to get it to run. I had some entry points, and this was the way I entered back into physics, which is the kind of work I do now.


Konstantinos Karagiannis:

That makes sense. Back in the early days of computing, in the early days of hacking, you had to understand way down to the lowest level. When you get to quantum now, a lot of people stay pretty much above the abstraction layer — Qiskit or something like that — and that’s as deep as they want to go.

Like I said, this is a unique episode, and we’ll get to why. On to OpenQuantum, and I already could hear some of the seeds of OpenQuantum in what you said. To quote from the project, it’s “a complete open source blueprint for an incredibly low-cost hardware and software platform for controlling and manipulating ultracold atoms.” You heard we’re manipulating atoms here. I wasn’t kidding that this is a unique episode. Tell us what this mission statement means and how you got the project started.


Max Shirokawa:

It means a lot of things to me. One thing that is important to me is scientific accessibility. There are a lot of places in the world that don’t have access to a lot of scientific opportunities, and there are brilliant people everywhere. The limiting factor to getting these people driving impact in the world is showing them what’s possible.

For me, on a much smaller scale, I didn’t realise what you could do with the tools. The tools that are available in engineering have been driven down so low in cost these days, and a lot of people don’t realise that. Three-D printers, which are something I took a lot of inspiration from, 10, 15 years ago, were this research tool was pretty much only in academia or research labs in industry, and absurdly expensive.

Over the last 12 years, this incredible open source movement has driven down the price so low that you can get an Ender 3 off of Amazon right now for $150 that is probably printing at about the same quality you would get from a $100,000 machine 15 years ago. You can see the effects of that. There are all these websites, all these communities all over the internet of people building their own 3D printers, collaborating, sharing files, learning about materials science, about CAD, about industrial design — all these opportunities they’re unlocking just by having the resources available to them to learn more, to experiment, to play.

As far as quantum computing goes, it’s a little more esoteric, a little more niche, than 3D printing, of course, but it’s something everyone who’s interested in physics will probably interact with at least once in their life because it’s such a fascinating idea. Quantum mechanics in general, the way our world works at the nanoscopic scale, most people who have any philosophical lean like to think about it. The idea that we’re at a point in time where we can individually engineer these systems and manipulate individual quanta of matter and observe these phenomena is fascinating.

It would inspire a whole new generation of young students to do science, to do physics, if they knew how accessible it could be. The goal of OpenQuantum is to make that possible. The specific means of that is, well, I’m building this device. It allows the trapping of cold neutral atoms. The thing about atoms is that they are individual quanta of matter. They’ve got quantised energy states. If you can control individual atoms, you now have these objects for which you can manipulate their states at will. You can observe these quantum phenomena.

The way it got started is, I was taking a course in atomic physics, and there was no final exam. It was a project-based course. Most people normally did a literature review, but I was very hands-on. I asked if I could build something. I built this ion trap. CERN has this cute open source blueprint for a macroscopic ion trap. I put that together and found it unsatisfying because it took me about a couple of days. I asked, “What if I built the whole magneto-optical trap?”

It’s out there on the internet. It’s a cool tool. The professor’s, like, “Sure — why not? Give it a try.” This was during the first year of COVID-19. I was at home in my New York apartment with a 3D printer, and I had learned how to build a lot of this stuff already working in this lab that I was doing separately. I started seeing what I could hack together. The results were immediately pretty promising. I had access to a lot of tools through the lab, which helped me speed up my process. But it seemed like something that could be made very accessible, and that was very exciting to me. That’s how this all got started.


Konstantinos Karagiannis:

Instead of baking bread, you were creating something pretty cool that everyone could try here for themselves. To clarify, this isn’t building a quantum computer at home, even though that would be incredibly cool — like going back to the Altair machine and Popular Electronics years ago. But you’re building a platform. Can you explain the difference?


Max Shirokawa:

Neutral atoms are individual quanta of matter, which means you can use them to observe, manipulate, produce quantum effects. That can take many forms. In fact, quantum computers, they get a lot of hype. But in terms of practicality, I don’t think they’re a good example of what’s possible with quantum engineering in the present.

Probably the best example of that that has the most practical impact worldwide is atomic clocks. Atomic clocks are based on quantum effects, and that’s how we synchronise time on missions that have relativistic implications, like on satellites. Most satellites have an atomic clock on board to keep time. These clocks are incredibly stable. Also, the time standard of the world — the way the second is defined — is based on an atomic clock that I believe lives in Colorado at NIST.

This is all based on the ability to manipulate atoms. That’s what I want to provide. It’s a platform that allows you to cool down atoms to such low temperatures, collective temperatures of about 10 to 100 microkelvin, which means their kinetic energy is low enough that you can start to manipulate them at will. It’s a platform for a wide variety of quantum experiments — more quantum experiments than there are existing magneto-optical traps in the world, which is also a motivating factor for me. One of those applications is a quantum computer, but you can do many other things with it.


Konstantinos Karagiannis:

Some of our listeners have heard of QuEra and Atom Computing right on this show. But give us a quick refresher on neutral-atom quantum computing as being based on the same idea.


Max Shirokawa:

The way neutral-atom quantum computing works is, you’ve got these cold neutral atoms. As a disclaimer, I’m not a theorist, so this is going to be very high-level, but you can encode logically, on paper, the ground state of those atoms as a logical zero, and then you pick some excited state to encode as a one. We picked a very high-level excited state, a Rydberg state. It’s also occasionally called Rydberg atom quantum computing.

What this allows you to do is purely optically use lasers to modulate these atoms between these two states, from the ground state into the excited state, or superposition of those two states, and then through more sophisticated optical operations, you can also do things like imply quantum gates, like a Hadamard gate, if you want to generate a Bell state. This is all handled optically. You’re taking the energy states of these individual atoms and using optical control to put them into specific quantum states you map onto some meaning at the software level, at the logical level, so you can run a quantum algorithm.


Konstantinos Karagiannis:

And the current state of the art on that, Atom Computing just announced a thousand qubits, and I’m assuming QR was going to be behind, because that’s a similar approach. To get a sense of how the technology is advancing, what are the specs of your device? What are people going to be realistically playing with here?


Max Shirokawa:

I’m building not necessarily a quantum computer, but more of an atom trap. For these companies, they’ll start with an atom trap that probably has about 100 million cold atoms that are floating around in this little ball in space. They’re still moving very slowly, but they’re moving nonetheless. To generate the basis for their computer, they have this array of optical traps called optical tweezers. This is a technique that won the Nobel Prize in 2018 — fairly recently — and once they’ve trapped these atoms in this optical trap, these atoms are basically not moving at all.

Each of those trapped atoms is a qubit. When Atom Computing says they have a thousand qubits, they mean they have this floating grid of atoms trapped in optical tweezers. Now they can manipulate them again at will. My device will allow you to trap, to cool down, about 10 million to 100 million rubidium atoms. From that point, if you had something like a spatial light modulator and a trapping laser of sufficient power, you could probably trap some dozens of atoms if you make a high-fidelity optical-tweezer array. That is something I hope someone will build on top of the platform. But what I’ll give you are the cold atoms you can then play with.


Konstantinos Karagiannis:

All this evolved from magneto-optical traps, and that’s like a 1997 technology. The project states that you can build your own completed system here for $8,000 or so. Tell us how you got that cost down. You already mentioned 3D printing.


Max Shirokawa:

Three-D printing — additive manufacturing — is a huge component of that. The motivation there was, very few universities — very few institutions in general around the world — have access to one of these machines. Even within America, it’s a pretty small fraction of educational institutions that have a teaching lab for quantum engineering. Most teaching labs for physics are things like kinematics, where they’ve got students rolling balls downhill, and in my opinion, that just makes fewer students want to do physics at all, whereas a teaching lab in quantum is inspiring because it shows you all the potential of what the future could look like.

But the issue is that now, the equipment is expensive. There’s a company called Infleqtion, formerly known as ColdQuanta, and they offer almost prebuilt magneto-optical traps. But I believe the cost is about $55,000, which is not a terrible amount of money, but for a community college or even a smaller private college, it’s going to be inaccessible. It’s hard to justify. To drive bigger, broader impact, to make it more accessible, probably the most important thing to do is to get the cost down.

As I learned as I was building the components necessary to establish the optical trap I was working on at this academic lab, it didn’t seem like there was any good reason it was so expensive. Again, there’s economies of scale. There’s a lot of labor involved that drives these prices up.

But that’s where I see something like 3D printing, digital fabrication in general, coming into play — digital fabrication being the concept of, now there’s no skilled labor involved. I can design something in a CAD programme, convert it into a mesh in STL, put it on the internet and give someone some very brief printing instructions, and they can have almost an exact replica — printer quality — of what I have on my desk, and they don’t need to know anything specific about the manufacturing process. To me, that’s an exciting idea. I see that as a way to allow people to get on board with this without having to have them learn all the fabrication techniques I had to become familiar with to begin interacting with this.

A lot of the components in a commercial or academic magneto-optical trap or any optical device will be machined from aluminum, sometimes steel, sometimes something like Invar. That also drives the price up. There are a lot of reasons for that in terms of stability. But if you want something for a teaching lab, you’re not going to be building a company off this, necessarily.

Polymers are more than enough. They’re slept on, in my opinion. People disregard them because it’s like it’s plastic — it’s a toy. But for establishing a locked laser for some number of hours, sufficient for an undergrad experiment, it’s plenty. I can take all these high costs — individually, they’re not incredibly high costs, but again, you need a lot of them to build a device like this, and it adds up quite quickly. I can design bespoke equipment for the specific optical elements needed for this device and throw the designs online, and anyone around the world can download it and print it for the cost of the filament — the plastic — which is cents per gram. That’s one of the primary ways I’m able to drive the cost down so much.


Konstantinos Karagiannis:

Duty cycle is a big thing. Depending on the material you choose, what’s it going to be exposed to, what kind of rigorous environment? That makes perfect sense. To give our listeners a sense, can we go through the key components of the system so they can visualise what this will look like when they decide to devote a whole room to it?


Max Shirokawa:

Hopefully, you won’t need a whole room. You can fit it on a desk at this point. The nice thing about the magneto-optical trap from the engineering teaching perspective is that it’s a pretty modular device in the sense that there are two main subsystems: an optical system and a vacuum system. Then you can break both of those down into further modular systems.

I’ll start with the optical side because it’s the more complex side. That’s also where at least someone interested in science, in quantum, will have the most fun, will get the most out of it. How do we cool down atoms to ultracold temperatures? Without going into too much detail, it’s about accessing their atomic transitions in a precise manner. Atoms move between energy states by absorbing and emitting specific amounts of energy. That’s why it’s quantised: It’s discrete. When you get too far away from these very specific values of energy, they won’t interact with whatever form of energy that is in general.

What we know from quantum optics is that light is quantised. The quanta of light is a photon, and each photon has some energy associated with its frequency or its wavelength. If we can get a laser such that the frequency of the photons that laser is emitting is of almost the exact frequency of the transitions of these atoms, we can force these atoms to interact with that light. For that, you need a laser that is very frequency-stable, because if the energy of the photons is fluctuating too much, you’re not going to be able to do this in a repeatable manner.

Precision is the key here. The first component is this frequency-stable laser. Every photon you’re emitting from this laser is roughly the same energy. I won’t go into too much detail about how that works, but please feel free to check out my website if you’re interested in the technical details.


Konstantinos Karagiannis:

That’ll be in the show notes.


Max Shirokawa:

Great. The next component is, now you’ve got something that is precise. Every photon is roughly the same energy, but now we also need accuracy because we need to be at the energy of the atomic transition of one of these atoms. This involves some spectroscopy, understanding the spectrum of atoms. The spectrum of an atom is these energy levels.

The way we do this is with a reference cell of atomic vapor of the same elements you want to trap. In my case, the rubidium reference cell I first started using when I embarked on this project was from an old Soviet atomic clock, which is cool. It still sits on my desk. As you pass light through this vapor cell, if your photon energy is close enough to the transition energy, it will start to glow. You’ll start to see it emit light because the atoms within that cell are absorbing those photons you’re putting through it. The closer you get to the resonant transition frequency of the atoms, the brighter it will get, because more and more of those atoms will absorb it.

You can measure this pretty simply with a photon detector. In fact, we put a photon detector right behind the vapor cell. As we get closer to this resonant transition frequency, you see less of a response at the photodiode. That’s because the atoms are absorbing that light, and that light is not making it into the photodiode. That’s a way to start to establish that you have a laser that is now both precise and accurate in terms of frequency. At a high level, that’s the primary optical system of the magneto-optical trap.

A lot of the cost associated with that is the optomechanics — the way of manipulating light mechanically. This means precise kinematic mounts that allow you to change the beam direction — mounting beam splitters in a precise way — because alignment is one of the most critical parts of any optical operation. People spend a lot of money figuring out how to keep that repeatable and stable. But I’ve designed the system from the ground up to use 3D-printed components. Everything is designed for my system, rather than being a bread-board-style optical setup. It makes it both compact and much lower-cost than using off- the-shelf commercial machined components.


Konstantinos Karagiannis:

You could still get the accuracy for lasers with that kind of 3D-printing approach.


Max Shirokawa:

Absolutely. When I started, it was a lot of stuff I had machined or purchased or got off eBay. Over the last couple of years, I’ve been replacing them one by one with a 3D-printed component until almost everything at this point is printed. The other subsystem that is involved is the vacuum system. You need an ultra high-vacuum chamber for a magneto-optical trap. This is, for a number of reasons, primarily because, for instance, the molecules in air, they have quite high energy. Even if you were able to cool down a cloud of atomic vapor like rubidium vapor to ultracold temperatures without a vacuum, it would quickly regain high energy by colliding with all these ambient molecules.

Another issue is oxidisation. You don’t want to form any oxides. And spatially, you want to confine it within some region of space — not have it float around whatever room you’re in. Ultra high-vacuum is required, probably roughly about 10-10 torr, which, even for a hobby vacuum, is quite low-pressure. This is where a majority of the cost comes. This is at this point where a majority of my efforts go in terms of lowering the bill of materials.

But at least mechanistically, it’s quite simple. You have this ultra high-vacuum chamber, and there is a series of pumps: a roughing pump, a high-vacuum pump like a turbomolecular pump, and then something to get you in that ultra-high regime, like an ion-getter pump — three stages of pumping.

Let’s assume you have your high-vacuum chamber. Now all you need is a source of atomic vapor. The easiest way to do this is to have a little capsule of rubidium. You attach this to some electrical feed-through — some way of putting currents into your vacuum chamber. As you put current through it, some of those atoms will volatilise, and you’ll start getting atoms in the vapor phase in your chamber. It’ll release some atoms from the solid rubidium crystal in your chamber, and you’ll have some floating atoms you can trap. Again, we’re only going for 10 million to 100 million, which maybe sounds like a lot, but when you think that 1 mole of some amount of atoms is 1023, 10 million to 100 million is not that many.


Konstantinos Karagiannis:

And there’s variance probably in the price here because you don’t know if you’re going to find a deal on eBay on the vacuum or something like that. It could change.


Max Shirokawa:

That’s correct. That’s the big, open problem for me if I wanted to get this as low-cost as possible so as many people as possible can have it. This $8,000 price tag, it includes commercial, off-the-shelf vacuum chambers from websites like Lesker or Accu-Glass Products. But I want to get it even lower. As you mentioned, the vacuum chambers I started playing with came from eBay. My vacuum chamber, it’s probably less than $1,000 because you can find great vacuum components on eBay. Their entire internet community is dedicated to people hacking the stuff together at home, with some interesting science going on across a huge variety of fields.


Konstantinos Karagiannis:

That price could be prohibitive for some individuals. If they’re thinking $8,000, they might not want to do it. But groups and organisations, like you’ve hinted at, colleges and things — have any told you they’re building it or succeeded and sent you photos, that kind of thing? Do you know of any completed ones?


Max Shirokawa:

No one has completed anything yet. Part of that is on me for not making it easier to assemble. I’ve had this working in various forms over the years, but nothing as stable as I want entirely made of my design.

I’m talking to a lot of educational institutions at the moment. Ben Varco, who was also at Quantum Village at DEF CON, has some of his master’s students working on the optics side trying to replicate my designs. I’m in talks with some people at Sandia who think it could have some interesting educational impact. The University of Chicago is interested in my designs. They have an incredible quantum course, but they also are aligned with my mission of making it more accessible to people who are not in an institution that has that much money and resources. But I’m still in the process of trying to gather interest from educational institutions and onboard people about not just getting the equipment, because the equipment is only one part of it.

In fact, from the educational perspective, it might be the less important part relative to a curriculum because it doesn’t matter if you have the equipment if there isn’t a good way to teach people how to use it. That’s the second component of OpenQuantum: not just providing people with files and saying, “Go have fun — go figure it out,” but also providing educational resources. I’m still figuring out whether that’s going to be YouTube videos or lecture notes that would allow someone to self-study this, even maybe perhaps without building the equipment, just reading through the documentation. I know I’ve learned a lot about hardware without ever having to build the hardware I’m reading about — just learning how it’s put together and what goes into making it. In fact, those are my favorite articles in general: the ones that tell you exactly what you need to do to make this happen.

For a teacher that maybe doesn’t have an incredibly deep understanding of atomic physics, or a teacher that hasn’t interacted that much with a CAD programme or 3D-printed a lot of things, how can I provide a resource to them that still lets them feel confident in teaching a classroom of kids how to use this device? That’s just as important because these institutions lack perhaps not only people or equipment but also the people with the domain knowledge to run such equipment, even if it did exist. That’s my goal going into the next year.


Konstantinos Karagiannis:

It’s an amazing idea because years ago, people would read Popular Electronics, and they didn’t necessarily build every single circuit that was in print, but they would read the description of how the circuit works, you have this timer circuit, you have this, etc. You would learn even for the ones you weren’t building. Some folks might just learn by reading on your site and seeing the documentation, and some are going to learn by doing, which is amazing.

We’re at that stage now, finally, that people can experiment with the hardware side too. That’s why I thought this was such a cool project. Just to be clear, I’m sure some listeners are imagining they have to have a clean room or something or some kind of bunny suit to operate this, like in Intel commercials. But I imagine it’s not that kind of environment.


Max Shirokawa:

The cleaner your room, the better. The quality of the thing you produce is going to be a function of the care you take when you assemble it. There is some threshold below which you will not be able to establish any spectroscopic signal. But definitely, you don’t need a clean room — nothing ISO-graded. I certainly didn’t do it like that. But you do have to take some care in assembling it, and that will of course all be in the instructions and the educational manuals I want to put out.

That’s an important part of this whole project, and the goal of the project as a whole is — well, a lot of these things are arcane because maybe you have some intuition that these things need to be clean but you don’t know what that means. If that’s you, you’re not alone because that’s almost anybody who hasn’t held a research position or a position in industry in this field.

That’s a big problem for the industry, because that means anytime someone comes in, like a first-year Ph.D. student or even sometimes a postdoc who came from a different field or someone gets a job in one of these companies like QuEra, like Atom, or even a different quantum computing company — because all the technologies share a lot of commonalities — they have to spend two to three months getting up to speed, learning how all these technologies work, learning all the procedures, learning what is and what isn’t necessary. This slows down progress.

Nobody likes dealing with this. People want to learn these things. They just don’t know how. The people running these companies don’t want to have to wait a quarter of a year for people to have all the domain knowledge to not screw up an experiment, to be able to hit the ground running.

That’s one of the big needs on the educational side of this industry this project can at least in some part fill, where now, when you have a new student who wants to join a lab or a new employee who doesn’t have a Ph.D., they’ve had an opportunity to at least see, to interact a little bit with, all the constituent technologies that go into making this industry possible as a whole. They have a little more respect for what goes on. They have more understanding. They can do more from day one. That’s going to be critical if this industry is ever to succeed commercially —to have a workforce that doesn’t need to be educated from scratch every time.


Konstantinos Karagiannis:

They’re learning and they’re building. But in the end, they end up with this device — and it’s like we’ve said: It’s not quite a quantum computer. You hinted that you envision some next level of project that can be built on top so you end up with something. There are rumored to be coming things for like, $5,000, a tiny little two-qubit machine or something sitting on your desk. I don’t know how useful that’ll be. Do you envision an extra layer on top that makes it, like, “In the end, I can run a very simple gate”?


Max Shirokawa:

I love those little two-qubit machines. They’re illustrative and informative, but they’re not extensible. You’ll have two qubits, and for the majority of those, that’s all you’ll ever get, whereas with my platform, you’ve got 100 million cold atoms. You can do a lot of things with it — a lot of things outside of quantum computing — if you’re just interested in quantum engineering in general.

Some of the basic things you can do, for instance, in an undergraduate lab, are measuring the rate of Raman scattering or characterising the line width of an atomic transition. These are things that will let you experimentally verify the laws of physics, which, to me, is satisfying because I’m taught all this stuff. I learned a lot of linear algebra. I learned a lot of quantum mechanics, a lot of atomic physics. For all I know, it could all be fake, except that once you put it together and you can compute a measurement that is within some epsilon of what you were taught in theory, now you’re, like, “This is very satisfying. We can verify that all these things are true. I can verify these things for myself. I don’t even have to rely on someone who told me he did it.”

To your point earlier, I don’t expect a lot of hobbyists, at least for the time being, to try to put one of this together. it is a bit cost-prohibitive, and there’s probably low motivation to spend that much money on something like this. Potentially, someone entrepreneurial could see a big gap in something like quantum sensing. As an aside, there are some interesting projects in quantum sensing. There’s a group at Berkeley that is using atom traps to do what they call quantum cartography, which is using atoms as a gravimetric sensor to scan the topography of terrain.


Konstantinos Karagiannis:

We talked with Infleqtion about this idea — detecting, with a truck-sized device, giant openings in the Earth below, and stuff like that.


Max Shirokawa:

That stuff is amazing. It’s so practical. We’re taking something that used to be so arcane and making it into something we can utilise to improve our control over the world. It’s incredible to me.

Another application I want to highlight is the magnetometer. There are a lot of people working on atomic magnetometry, which is a way of measuring magnetic fields with atoms. Atoms are so sensitive to these external fields that you can use them to get more precise measurements than you could do with anything classical, to my understanding. Perhaps someone entrepreneurial could find some gap in one of these fields and use a platform as a starting point for experiments to de-risk their idea before they potentially raise venture capital.
On the educational side, what I hope that comes of this is, people start building experiments on top of it. For instance, one of the first things scientists did with the magneto-optical trap was make a Bose-Einstein condensate, which is one of the most exotic states of matter. I don’t think that would be out of reach for my device. I don’t necessarily plan on doing it, but it would make me happy if someone did.

Then, of course, quantum computing — what we’re all here for. Maybe you’re not competing with Atom Computing. You don’t have a 1,000-qubit array. But I don’t see why with someone who does something clever with some crossed acousto-optic deflectors wouldn’t be able to trap some handful of atoms with optical tweezers and then run a deterministic excitation to a Rydberg state with that.
Again, this is where my theoretical understanding of the field probably hits the limit, but the equipment I want to provide is not too far off, I hope, from lab-scale equipment. If you wanted to go much further with it, I want to provide at least the platform for someone to try to do that.


Konstantinos Karagiannis:

That’s wild. Who knows — maybe a year from now, we’ll be talking about the successes in building the first home-brewed quantum computer. It’s possible.

This has been a fantastic conversation. The holiday season is approaching. People have some downtime. Who knows — there might be an uptick in building.

Before I let you go, what are ways folks can get involved if they’re not ready to build — other ways they can get involved with this project?


Max Shirokawa:

At a logistical level, I would love to be put in touch with any educators, any institutions, that might be interested in at least having a conversation about what this could look like. That’s how I can drive the most impact with this. If you have strong opinions about education, about atomic physics, and you have input feedback on my curriculum, my educational resources, I would love to hear it.

If you’re not ready to build, well, maybe you should be, and we can talk about that because these tools are so much more accessible even than I thought they were three or four years ago. All it took was having those tools in place and having a little nudge to see what’s possible with them to get started. Everything has snowballed from there. If anybody’s on the fence about getting their hands dirty, buying a 3D printer, playing with some CAD models, implementing that experiment they’ve been thinking about for several months, I would say go for it, and you can reach out to me. I’d love to chat.


Konstantinos Karagiannis:

Yeah, there’s going to be some great eBay vacuum-device sniping going on over the next few weeks as we’re trying to grab a device from each other. Max, thanks, again, so much. Good luck with the project. I’m happy to be able to spread the word here, and let’s see what happens.


Max Shirokawa:

Thanks again for having me. I really appreciate it.


Konstantinos Karagiannis:

Now, it’s time for Coherence, the quantum executive summary, where I take a moment to highlight some of the business impacts we discussed today in case things got too nerdy at times. Let’s recap.

OpenQuantum is a unique project that provides blueprints for building an ultracold-atom platform. Whether you’re a serious enthusiast or perhaps looking for a project for a class activity in an educational institution, OpenQuantum’s approach is unlike anything else out there in this price range. A combination of 3D printing and carefully chosen components make building the platform possible for $8,000 or less. The variability in price is due to you maybe getting lucky with vacuum chambers and other parts on eBay. The platform allows you to cool down 10 million to 100 million neutral rubidium atoms to temperatures at sub-100 microkelvin.

While it’s not a fully functioning quantum computer and you won’t be running gates, the platform has a potential to be built on in the future. For example, adding a high-fidelity optical-tweezer array could let you use some of the atoms as qubits one day. It will be exciting to see if a community forms around such expansion in the future.

For now, the device will fit on a table or desk. The whole approach has a similar feel to what it was like to build computer kits in the 1970s, and look how that turned out for PC system builders eventually. If you’re interested in more of the engineering side of quantum systems, this project is definitely worth checking out.

That does it for this episode. Thanks to Max Shirokawa for joining to discuss OpenQuantum, and thank you for listening. If you enjoyed the show, please subscribe to Protiviti’s The Post-Quantum World, and leave a review to help others find us. Be sure to follow me on all socials @KonstantHacker. You’ll find links there to what we’re doing in Quantum Computing Services at Protiviti. You can also DM me questions or suggestions for what you’d like to hear on the show. For more information on our quantum services, check out, or follow Protiviti Tech on Twitter and LinkedIn. Until next time, be kind, and stay quantum-curious.