Colorado’s newest startup incubator recently opened its doors on the campus of the state’s oldest public institution of higher learning.
Atom Computing's Quantum Tech and the Story Behind It
John Himes
November 7, 2023
What’s smaller than a grain of sand, is as cold as the void of outer space, and uses lasers to control individual atoms? Cutting-edge quantum processors in Boulder, Colorado.
Atom Computing’s quantum technology doesn’t just move atoms around. It encodes information into them, puts atoms into liminal quantum states that involve entanglement and superposition, and then runs algorithms that solve complex problems that even today’s largest supercomputers would struggle to churn through.
Quantum computing (or QC, which also stands for quantum computer) may be a nascent technology that’s still a few years from mainstream commercial viability, but you’d never be able to tell if you just looked at Colorado’s booming QC industry. For several reasons, quantum has taken root on the Front Range.
We’re now seeing an effect that’s similar to what the electrical transistor did for Silicon Valley. Quantum technology companies compete and collaborate, while upstream high-tech manufacturers, service providers, and educational institutions grow in tandem with increasing demand.
This gravitational pull was one of the main reasons that Atom Computing, a company founded in Berkeley in 2018, decided to set up shop in Boulder for their second location, the one where they would transition from academic prototype to commercially viable B2B tech product.
Of course, the fact that Dr. Ben Bloom, the company’s founder and CTO, completed his PhD at the University of Colorado (CU) Boulder in 2014 was part of the calculus. He understood that Colorado is the place to be for quantum tech development.
The quantum computing endgame
First we need to understand why investors are pouring $2.5B in annual investment into quantum startups in the first place. And that’s just in the US—and it doesn’t even account for the unfathomable money already spent by tech giants like IBM, Google, and Intel.
These juggernauts are competing with startups like Atom Computing to build scalable, reliable quantum computers.
There’s a ton of potential upside. And if there’s one thing we’ve learned from how far scientists and engineers have already come with this difficult and often counterintuitive technology, it’s that many of these theories will inevitably be engineered into practical machinery.
“One of the killer applications for quantum computing is combinatorial optimization problems,” says Kortny Rolston-Duce, Atom Computing’s Director of Marketing and Communications. These problems include tons of variables; real life applications include “managing supply chains, devising efficient transportation routes, and balancing energy load across the grid,” she explains.
Right now, these problems are mostly solved by people’s experience alongside a bit of modeling, but the problems are just too big for even the world’s biggest and baddest of supercomputers.
It comes down to the fundamentals of what’s happening in the hardware: since an electrical transistor can only be an individualized instance of “yes” or “no,” the computer has to brute force its way through every possible solution. That’s the problem with bits.
When problems become too big, like the intense complexity of managing every single power supplier and consumer—down to the smallest toaster—well, it’s just not possible for classical computers to work through them in a time frame that fits within our own human lifetimes.
Quantum flips that paradigm. Qubits, the fundamental building blocks of a QC, get around both the individualization and the strict dichotomy of classical bits by leveraging the quantum mechanical properties of entanglement and superposition, respectively.
What's entanglement and superposition got to do with it?
Entanglement, which Einstein famously called “spooky action at a distance,” is when two things, in this case the nuclear spin of two atoms, are interdependent in some fundamental ontological way.
To be honest, nobody really understands why. Something happens to atom A, and atom B will be affected, no matter where it is or what else is going on. This kind of faster-than-light communication overcomes the individualization problem of electrical transistors.
The second part, superposition, solves the dichotomy issue, the unrelenting “yes” or “no” of classical bits. Superposition allows matter in a quantum state to be “kinda yes, kinda no; ask me later.”
It’s probability manifest. Think of it like making bets, like game theory. But instead of determining the likelihood of an event in the future, it’s a state of matter right now.
f you bet on a horse race, the bookie will give you the odds, the likelihood of your horse winning. Once the race is over, you have your answer—the probability has collapsed into something finite.
The same thing happens with qubits in superposition when they’re measured. They can physically be “kinda yes, kinda no” while they’re in a quantum state and working through a problem, but as soon as you ask the computer for a solution, the probability collapses into a single state: yes or no.
We then make this calculation over and over again. Sometimes we get different results. But when we run it enough times, we get the most likely answer to our problem, which, it turns out, is the answer we’re looking for.
Let’s say we want to find the ace of spades in a standard deck of cards. Without any additional knowledge, a classical computer will have to randomly pick cards from the deck, and we’ll get lucky once in 52 times. That’s not very efficient.
In a quantum computer, this problem is very different because we can design a quantum circuit such that the probabilities of all the wrong answers (all non-ace-of-spades cards) cancel out, leaving us with only the ace of spades as the answer. As a result we will end up with the ace of spades much more frequently than once in 52 times.
Putting quantum to use
In a nutshell, this is why quantum computers are so good at combinatorial optimization.
This capacity makes QC attractive for a lot of practical applications. One well-documented one is breaking classical encryption; it turns out that figuring out large prime numbers, which is the basis of modern cryptography, is trivial when you don’t have to brute force your way through every possibility. Materials science and drug discovery are other oft-cited applications for similar reasons.
But we still haven’t touched on the biggest one. At least in my opinion, the main reason so much money is flowing into QC development comes down to two letters: A and I.
You think today’s machine learning models are powerful? Right now, a single request requires tons of calculations, and training a model requires even more.
When QCs provide the power of combinatorial optimization, I would not be shocked to witness an intelligence explosion.
And who knows? One of the killer apps for classical computers and the internet turned out to be cat videos. Maybe for the quantum computer, it will be Schrodinger’s cat videos.
What is the atomic array of neutral atoms modality in quantum computing?
Atom Computing’s unique hardware uses light to bring individual atoms to a standstill, arrange them into a grid, and then manipulate them to run computations.
To turn atoms into qubits, Atom Computing releases a stream of alkaline earth metal atoms into a vacuum chamber. Then, rather than using cryogenic cooling like many other QCs, they use strong magnetic fields and lasers to bring them asymptotically close to a complete stop: around 100 nanokelvins, colder than interstellar space.
In this way, they create “cold” neutral atoms. The vacuum chamber protects these atoms from outside interference, but the ambient temperature of the chamber itself remains at room temperature.
They then employ optical tweezers, which use “a tightly focused laser to generate a trapping force that can capture and move” the atoms into an evenly spaced 2D array. These lasers are calibrated to perfectly match the energy potential of the atoms; the atoms “want” to sit inside the laser, and this technique can even hold up atoms vertically against gravity.
Now that the atoms are in a grid, we can start running quantum circuits.
Qubits are born
They shoot another beam at the atom from the side to manipulate its nuclear spin. A combination of proprietary control electronics and radio frequency encoding enables them to use the interaction between the lasers and the atoms to form logic gates.
Essentially, just like electricity flowing through a transistor represents a gate (or bit) in a classical computer, the laser-controlled nuclear spin of the atom is a qubit.
Unlike a classical bit, which can only be either 1 or 0 at any given time, the atom’s nuclear spin, and thereby the qubit, can exist in a quantum state that can be both 1 and 0 at the same time. As we discussed in the previous section, this superposition is what gives qubits the advantage of probability made manifest.
Atom Computing achieves the other aspect of the quantum state, entanglement, by shining an ultraviolet ray across the array. This pulse puts an atom into a highly energized Rydberg state that entwines it with a neighboring atom. The result is that any change in one atom’s spin will instantaneously affect its entangled partner.
These are called two-qubit gates, and they’re largely responsible for the QC’s ability to handle immense complexity. While the individual gates of classical bits scale linearly, the computing power of two-qubit gates scales exponentially as we add more gates.
Finally, after running the quantum circuit, Atom Computing uses a camera to read out the results. Remember, quantum computers are probability machines. So each time a circuit is read out, the qubits provide a single answer, but the circuit needs to be rerun multiple times to build a histogram, a distribution of all possible outcomes sorted by their relative likelihood. To gain more confidence that the most likely answer is the best one, Atom Computing generally runs a single circuit many times.
Don’t forget that all of this happens within a space that’s as tiny as a grain of sand.
Curious readers may choose to go deeper with Atom Computing’s technical white paper.
What are the advantages of atomic array quantum computing?
This type of quantum computer is one of several architectures competing to be the breakthrough technology that brings quantum into the mainstream. Other modalities include superconducting, ion trap, and different neutral atom setups.
So why is Atom Computing betting on atomic arrays?
The first reason is scalability. “Using light as a medium lets us scale by many orders of magnitude,” explains Dr. Remy Notermans, Director of Strategic Planning.
The company recently announced a 1,000+ qubit quantum computer, making it the QC with the highest qubit count to date.
While other modalities, such as superconductive QCs, have to go through the immense challenge of fabricating silicon wafers to an incredibly high degree of precision, manipulating qubits with light is a bit simpler. It still requires complex optical engineering and special lenses, but it’s not as sensitive.
The other aspect of scalability comes from the topology of the array itself. Atom Computing’s prototype machine, which they call Phoenix, is a 10×10 grid, but they’re able to scale to much larger sizes with relative ease and within the same grain-sized footprint. They plan to eventually scale into a 3D array.
The second reason is fidelity. Fault-tolerance remains a major problem for QC tech in general, and improving error correction with fault-tolerant qubits is a goal for every quantum hardware developer.
One boon to the company’s atomic arrays is the fact that they use nuclear spin qubits instead of electron spin qubits. Since information is encoded in the nucleus, it’s shielded by the atom’s electron cloud. This means that the qubit can maintain a quantum state of superposition and entanglement—what’s known as coherence time—for longer than many other quantum modalities.
In 2018, Atom Computing set the world record for coherence time: 40 seconds. It may not sound like much, but compare that to IBM’s latest performance metric of 400 microseconds of coherence time. That’s 0.0004 seconds, five orders of magnitude less than Atom Computing’s coherence time.
The upshot is that atomic arrays are less sensitive to latency, which relieves some of the burden on the control electronics. This also means that quantum circuits have more time to run error correction code to achieve higher levels of fidelity.
The third advantage of atomic arrays is their ability to function at room temperature. While other QC architectures rely on sensitive, expensive, and complicated dilution refrigerators to cool the quantum processor to cryogenic temperatures, atomic arrays forgo this necessity. They make the atoms cold by trapping them in light instead.
Since the rest of the hardware remains at room temperature, this simplifies the operational aspects of running the quantum computer.
LIKE WHAT YOU’RE READING?
Get more, straight to your inbox.
Portrait of a quantum innovator
Dr. Bloom, the company’s current CTO, founded Atom Computing in 2018 after finishing a PhD program at CU Boulder that focused on optical atomic clocks. He saw the progress other companies were making on quantum technology, and he had an idea for a new paradigm: neutral atoms organized into an array.
Dr. Bloom and his cofounder, Dr. Jonathan King, built out the team and got to work on their prototype hardware, named Phoenix, in Berkeley. It wasn’t long before they had a proof-of-concept; they proved that their QC modality had the benefits of sustained coherence, scalability, and relative simplicity compared to others.
In September 2022, they opened a facility in Boulder as they set out to build their 2nd generation system and develop their technology into a commercially viable state. They rapidly expanded their headcount to over 70 employees, and more than 50 of them hold PhDs in fields ranging from quantum physics to electrical engineering.
Since then, they’ve built out their product on both the hardware and software sides. For instance, they’ve created modular optical tables that slot into their computers to change functionality, optimize for different tasks, and run comparative tests.
Compared to the tightly integrated prototype system, this is a big gain in terms of commercial viability and being able to iterate as they move forward. It’s the same story as classical computing: the invention of modular hardware was a true game changer.
“Modularity also helps simplify complex systems and divvy up complex tasks so that individuals and companies can focus their efforts productively on a manageable set of objectives,” says Carliss Baldwin, Professor of Business Administration at Harvard Business School.
Not just quantum hardware
Atom Computing is a hardware company, but software remains an ongoing challenge. They have made breakthroughs in some areas, such as developing a quantum circuit compiler that optimizes how variables are assigned to individual qubits to maximize topology efficiency for nearest-neighbor entanglement.
One main objective of their software R&D is to enable interoperability of quantum algorithm code. Just like how C changed classical computing by introducing a universal programming language that would operate on any machine—whereas previously every model had its own proprietary way of programming—quantum computing is on a similar trajectory.
To reach potential users where they are, Atom Computing’s developers are making sure that programs written in languages like Qiskit, the industry’s de facto quantum toolkit, can run on their hardware.
Then, of course, there’s the biggest challenge of all. “This is a brand new field,” says Rolston-Duce. “It hasn’t been built before. There’s no playbook or directions. It’s one thing to prove something in a lab, but engineering systems is very difficult.”
It’s no wonder they have so many bright minds working on it.
The road forward
The path to commercialization may not be easy, but they do have a vision. They recognize that mainstream QC adoption is still likely years away, but they want to provide a testbed that forward-looking companies can begin experimenting with to develop an understanding of how quantum computing works.
“Quantum computing is so counterintuitive that you need to experiment with it to understand the applications and possibilities,” explains Dr. Notermans. His advice for companies that see the potential and don’t want to be left behind: “Take the time now. Don’t wait.”
To this end, Atom Computing plans to launch a Quantum Computing as a Service (QCaaS) offering in the near future. They realize that it’s unrealistic to expect organizations except those with the deepest pockets to place a purchase order for their own quantum hardware.
Instead, they will provide cloud-based access to their hardware that enables users from around the world to run programs on the QCs in Boulder.
Finding inspiration on the shoulders of giants
The discovery of quantum theory was itself a giant leap and a testament to human ingenuity and progress. Harnessing the counterintuitive and difficult-to-control properties of the ultrasmall is an even bigger achievement.
A quantum computer is an engineering marvel that epitomizes the bleeding edge, so it’s not surprising that Justin Ging, Atom Computing’s Chief Product Officer, finds inspiration by standing on the shoulders of giants.
“We’ve put so many Nobel Prize–winning technologies together,” says Mr. Ging. “They’re all groundbreaking in their own right, and now we’re combining them to do something useful.”
He points to the winners of the 2022 Nobel Prize in Physics who explained quantum entanglement, which is fundamental to QC hardware. Another key enabler of Atom Computing’s hardware is optical tweezers, whose inventors won the 2018 Prize in Physics.
The list goes on: the winners of the 2012 Prize dealt with “ground-breaking experimental methods that enable measuring and manipulation of individual quantum systems.” The 2005 Prize was split into two equally relevant halves, one for “the quantum theory of optical coherence” and another for the development of “laser-based precision spectroscopy,” which makes it possible to construct the finely tuned lasers that the company uses to manipulate individual atoms.
For the Atom Computing team, the satisfaction lies in putting these breakthroughs to work and seeing the results for oneself. Dr. Notermans explains the thrill of “seeing a live stream of atoms being manipulated, seeing a quantum circuit happening in real time with your own eyes.”
While I was visiting their Boulder facility, I saw it too. Though the quantum computer itself is literally a black box, they had a screen in their office displaying the data coming out of it. In one corner, a live stream showed real-time video of individual atoms, neatly organized into a grid, being moved around.
It’s a powerful thing. Quantum physics and technology are infamously abstract, mysterious to the point of incomprehensibility to most everyone who doesn’t devote their life to pursuing them.
But that’s also what’s so amazing about tech. It’s pure application, a manifestation of the theoretical, of science that was painstakingly revealed over the centuries.
It’s something you can see with your own eyes.
Colorado’s quantum tech ecosystem hits critical mass
Colorado is the place to be for quantum tech.
The local quantum tech ecosystem has achieved “critical mass,” says Dr. Notermans, describing how Atom Computing has seen a “positive feedback loop” in which like begets like. As the number of quantum companies and adjacent industries grows, it attracts even more talent, investment, and innovation.
Simply put, because the quantum scene is so strong on the Front Range, people want to get in on the action, and this makes the scene even stronger.
Atom Computing is a case in point. They decided to open their second facility in Boulder rather than closer to their launch location of Berkeley because they saw the advantages of inhabiting this ecosystem.
And they’re not alone. Quantinuum, formerly known as Honeywell Quantum Solutions, took up residency in Broomfield in 2016. Infleqtion, formerly known as Cold Quanta and another neutral atom QC developer, has been working on quantum tech in Boulder since 2007. Just like in Silicon Valley, the combination of competition and collaboration pushes the tech forward.
In 2022, Daniel Strain wrote for CU Boulder Today that quantum startup companies “employ more than 1,000 people in the state and generate an estimated economic impact of around $400 million.” That figure has only grown as positive feedback spins the flywheel.
Partnership across the wider network
And this effect isn’t just localized to quantum tech companies. Partnerships throughout the state are essential. For instance, advanced optical equipment is a key ingredient for many quantum technologies, so the Colorado Photonics Industry Association and the companies it represents play into the equation just as much.
Optical suppliers have moved to Colorado, explains Dr. Notermans, so they can provide a stronger supply and service to their quantum customers. “There’s a strong personal component,” he says, likening the high-tech ecosystem to a small town.
There are also partnerships with other labs in the state. For example, Atom Computing recently teamed up with the National Renewable Energy Lab (NREL) in Golden to begin experimenting with QC for energy grid optimization.
They’re starting by figuring out the best way to draw power from different sources. And they’re using this opportunity to build the knowledge and experience to move forward once quantum technology reaches greater maturity.
This same story plays out in many public-private partnerships throughout the state. The National Institute of Standards and Technology (NIST) in Boulder is heavily involved in quantum applications like atomic clocks and cryptography. JILA, a joint institute of CU Boulder and NIST, also continues to advance the science behind quantum information, laser physics, and more.
And of course, the Department of Defense also has a major footprint in Colorado. Both the Air Force and Space Force play a prominent role, and many DARPA dollars continue to flow into high-tech capabilities like quantum computing.
Quantum education in Colorado
Educational opportunities—and the corresponding talent— are another driving force behind Colorado’s quantum tech ecosystem. Of course, CU Boulder continues to play an outsized role, but this institution is far from the only place that’s ramping up to supply the workforce necessary to grow the quantum sector.
The Colorado School of Mines likewise offers a quantum engineering program that they launched in 2020. Atom Computing has hosted Mines students in the past, and they found that the experience was just as enriching for their employees as it was for the students. These curious young minds brought a fresh perspective, asked new questions that they hadn’t considered, and helped to push the tech forward.
And it’s not just scientists and engineers that are in demand. Front Range Community College now offers an optics technology program that trains technicians on how to “fabricate, assemble, or install optical instruments,” work that is crucial for building QCs.
The future of quantum tech in Colorado
Colorado is currently competing for “tech hub designation” under the CHIPS and Science Act of 2022. This massive federal investment isn’t just about onshoring semiconductor fabrication; over $150M in annual investment will go to quantum technology in the coming years, and Colorado is making a strong case for a slice of that pie.
Additionally, the CHIPS Act appropriates $10B for NIST, which will certainly be a boon for the local quantum ecosystem, regardless of how the rest of the funding authorization pans out.
Private investment also continues to pour into the state. Atom Computing alone plans to invest $100M over the course of the coming years. This is part of a larger trend of venture-backed tech startups taking root in the state.
Colorado has the opportunity to become for quantum what Silicon Valley was for semiconductors. By attracting new players, developing talent, bringing in both public and private investment, and forming partnerships across adjacent industries and institutions, the Colorado tech ecosystem is incubating a quantum scene that’s unparalleled.
It’s an exciting time to build tech in Colorado. And this is only the beginning.
About the Colorado Tech Spotlight
The Colorado Tech Spotlight highlights local innovations and the stories behind them. The series explores how the Colorado tech ecosystem creates an environment that promotes technological progress.
It is produced by Dynamic Tech Media and written by John Himes.
Colorado Tech Insider
Stay up-to-date on the latest innovations and the stories behind them
Related Posts
Did you know that most black ink comes from fossil fuels? Living Ink is changing that by creating sustainable pigment out of algae.
Imagine a place where a physician can see patients in the morning, teach a class in the afternoon, go to the lab to check in on their research, then finish their day at their startup’s office—all without ever having to get in the car.