Alex Chernoguzov, the Chief Engineer of Commercial Products at Quantinuum, is helping to bring this programming platform to Quantinuum’s world-class quantum hardware.

“The more languages that support quantum, the better, because that opens up an opportunity for different software specialists to start programming in quantum environments,” Chernoguzov said. “We need to develop a new workforce that's educated on quantum information science topics and capable of generating new algorithms that can run on quantum computers.”

Tony Uttley, president and chief operating officer at Quantinuum, said platforms such as QODA are important for the company and the quantum computing industry. 

“At Quantinuum, our objective is to accelerate quantum computing’s utility to the world,” Uttley said. “By bringing forward additional tools like QODA, we expand the number of brilliant people aiming their talents at getting the most out of today’s quantum computers.”

Why we need QODA

Quantum computers speak a different language than classical machines. Also, the current landscape doesn’t have many effective quantum compilers to support interoperability with classical machines. The NVIDIA QODA platform aims to change that. Until recently, most quantum programming languages were based on Python because many scientists are familiar with it, Chernoguzov said.

“QODA adds quantum capabilities to C++ because this language is what's often used to program high performance computing machines,” he said. “Having a C++ dialect expands the possible languages that you can program quantum with.”

A classical-quantum bridge 

Chernoguzov said interoperability between classical and quantum systems was another core goal of this project. 

“Let’s say you have a hybrid program that has some classical parts and some quantum parts,” he said. “You compile the program. There is a classical piece that you can run on a CPU or a GPU, and there is a quantum piece that you need to send to a quantum computer. In a sense, you could look at it as a quantum processor acting as a co-processor for the other classical processors you need for your program. After completion, you gather everything together and do some more classical computations and repeat the process.”

Quantinuum’s H1 quantum machine will act as a quantum processor working in conjunction with larger classical systems. If a computational task has an element that could be solved more easily by a quantum architecture, this task can be passed off to H1 so researchers can solve quantum problems. This process will currently work in a similar fashion to other cloud-based services with programs submitted for execution over the cloud to H1. 

Quantinuum hardware and the NVIDIA QODA platform are bridging the gap between existing classical architectures and emerging quantum resources and using the strengths of each system to solve complex problems.

“Let’s say you want to model a complex chemical molecule. Atomic interactions are best handled by a quantum computer,” Chernoguzov said, “but directing the overall program flow to tell it what to model and how to model it is best done by the classical computers.” NVIDIA’s QODA platform helps reveal a world where these two ecosystems coexist and thrive together. 

Chernoguzov also explained the benefits of the Quantum Intermediate Representation (QIR) Alliance: a group of people and organizations who are committed to improving interoperability for quantum machines. This group’s work forms the basis for the hybrid approach that uses both classical and quantum machines.

“Interoperability in the quantum world is possible and the QIR is a good fit for that,” he said. “Quantum computers cannot do everything themselves, but classical compute is also clearly limited. We need both, and they need to work closely together to solve difficult problems that neither technology can solve on its own.”

When it comes to transporting ions, researchers at Quantinuum have turned a corner. Both literally and figuratively.

The Quantinuum team can now move two different types of ions through a junction in a surface trap, a tiny electrode-filled device at the heart of trapped ion quantum computers.

In a pre-print publication posted on arXiv, Quantinuum researchers outlined how they developed new waveforms that can guide a pair of ytterbium and barium ions through an intersection without the charged particles becoming overly excited or flying out of the trap.

The team tested the technique on a prototype trap with a grid-like architecture that Quantinuum has designed and microfabricated. This trap design will be a central part of future quantum computers such as the System Model H3.

This feat is an important breakthrough in the world of trapped ion quantum computing and for Quantinuum.

The ability to transport paired ions through a junction at the same time and intact is critical for scaling trapped ion systems. It’s also a longstanding technical challenge that trapped ion researchers in academia, government and industry have sought to solve for years.

“What Quantinuum has accomplished is a significant breakthrough for the field of trapped ion research and for our technology,” said Tony Uttley, president, and chief operating officer at Quantinuum. “This will enable us to deliver faster quantum computers with more qubits and fewer errors.”

Smooth transport of ions

Quantinuum’s technologies are based on the Quantum Charged Coupled Device (QCCD) architecture, a concept first introduced by the Ion Storage Group at the National Institute of Standards and Technology (NIST) in the early 2000s.

Like other trapped ion technologies, this architecture relies on traps to capture ions in electric fields - or wells. Gates are performed on small chains of ions, which can be reordered and reconfigured within the architecture, enabling all-to-all connectivity.

In Quantinuum’s System Model H1 technologies, each well contains an ytterbium ion and a barium ion. The ytterbium ion functions as a qubit while the barium is cooled with a laser to reduce the motions of both ions, a technique known as sympathetic cooling. This cooling makes it possible to maintain low error rates in quantum computing operations for long calculations.

The H1-1 and H1-2 machines currently use a trap with a simple geometry or design that resembles railroad tracks. Wells of ions are moved back and forth along these linear tracks and swapped as needed to run an algorithm.  

This linear design works well with fewer qubits. But it has limitations that make scaling difficult. Adding hundreds, much less thousands of qubits, would require the tracks to be much longer. It also would take more time to reposition and reset qubits.

To overcome these challenges, Quantinuum researchers have proposed moving to traps with more complex geometries. The System Model H2 will incorporate a racetrack-like design. The System Model H3 and beyond will use two-dimensional traps that resemble a city street grid with multiple railroad lines and intersections.

This design, however, also poses challenges. Getting those tracks to behave well at intersections is difficult and can jar ions and cause unwanted motion – especially those with different masses. It is somewhat like maneuvering a bullet train and allowing it to turn left or right at 90 degrees, or continue moving straight, without causing the cars to rock.

Quantinuum researchers were able to turn an ytterbium-barium ion pair around sharp corners with little motion. Until now, researchers envisioned having to separate paired ions and move them through junctions one a time, which would dramatically slow the operation. “To our knowledge, this is the first time any team has simultaneously moved two different species of ions through a junction in a surface trap,” said Dr. Cody Burton, a senior advanced physicist who worked on the project and lead author of the arXiv paper.

What’s next?

Researchers will continue to test and refine this new method.

Their goal is to expand from moving a single well to transporting several through multiple junctions at the same time. From there, they plan to incorporate this methodology into the System Model H3, which is expected to be the first Quantinuum quantum technology with the two-dimensional, grid-like trap.

“This new configuration will be key for scaling quantum computers in the hundreds, and then thousands, of high-fidelity qubits,” Uttley said. “While scaling, the qubits will maintain the high-quality characteristics such as low gate errors, long coherence times, and low cross-talk for which Quantinuum’s technologies are known.”

‍Mark Jackson is a man on a mission. As Quantinuum’s senior quantum evangelist, Mark’s job is to create awareness and understanding about quantum computing and its world-changing potential. Based in New York, Mark holds a Ph.D. in theoretical physics from Columbia University with a background in mathematical modeling and computational physics. In 2017 he joined Cambridge Quantum, which combined with Honeywell Quantum Solutions to form Quantinuum in 2021. He has an academic background and remains an adjunct faculty member at Singularity University. He sat down earlier this month to talk about his unique job and the future of quantum computing. 

Senior Quantum Evangelist is such a unique title. What does your job entail? 

A lot of my job is speaking at conferences, doing interviews, participating in podcasts, and posting on social media. I focus on creating awareness and excitement for quantum computing, letting people know what we do at Quantinuum, and educating them about the ways this amazing technology will help solve complex problems and improve people’s lives. 

Most people just don’t know much about quantum computing, or they have misunderstandings or reservations about the technology and its potential impact on society. 

Half the people don’t believe quantum computers really exist yet. They think it’s some sort of science fiction idea that we’ve cooked up and, if it happens at all, it’ll be 20 years from now. They just can’t believe we have these computers today. The other half think quantum computers are just really fast computers. They believe we can take all our existing software and run it on a quantum computer, and it will be a million times faster. Neither is true, and it’s my job to educate people about what quantum computers can actually do to make the world better. 

Over the past few years my role at Quantinuum has evolved a bit, and about a year ago they changed my title to “evangelist.” Technically, I’m now the “senior evangelist” because we recently added several other people to the team, which will help us do an even better job of spreading the word. 

How will the use of quantum computers benefit society?

We anticipate we’re only 3–5 years away from being able to do things on a quantum computer that are truly valuable to society. That time will pass very quickly, which is why we’re encouraging companies to work with us right now to develop projects so that in a few years, when technology catches up, they’ll be in a good position to take advantage of opportunities. 

The two nearest-term commercial applications for quantum computers are in chemistry and optimization, such as supply chain and logistics. 

In chemistry, we have known the equations for 100 years. If you give me a molecule, I know exactly what the molecule is made of — I know how many electrons, protons and neutrons are in it, and I know the equations governing all their interactions. But, solving those equations and actually figuring out the behavior of the molecule is very difficult because, as a molecule gets bigger, there are so many interactions that tracking them quickly overwhelms a conventional computer. Quantum computers are expected to one day solve these chemical equations easier and faster. 

For example, pharmaceutical companies could use this technology to design medicine. Right now, there is a lot of guesswork in developing a drug. Scientists can do a little preliminary work on a computer, but then they must synthesize a lot of trial drugs followed by testing on humans. 

Developing drugs this way is expensive, time consuming, and risky. In general, it takes about 10 years and $1 billion dollars to bring a drug to market. It would be ideal if scientists could do more work on a computer up front, which will save time and money and be less risky for patients. 

Additionally, quantum computing will be invaluable for the machine-learning industry. Artificial intelligence is used everywhere. Your Netflix recommendations use AI machine-learning, and while this may not be lifechanging, advanced autopilot technology on an airplane or in a driverless car will be. Quantum computers one day could have the power, speed, and capacity to take machine-learning to a whole new level. 

How did you end up working for Quantinuum? 

I started hearing about quantum computing in 2017 and thought it sounded amazing. This field of study didn’t even exist when I was a student. 

My background is in theoretical physics. For 15 years I worked in string theory and cosmology. Several years ago, I decided to leave academia and pursue other interests. I was very fortunate to be introduced to Ilyas Khan, founder of Cambridge Quantum and now CEO of Quantinuum, and he asked me to join the team about five years ago.

I was the first American hire at Cambridge Quantum, which was then a small start-up company with only about 30 people. The organization was comprised of all scientists until I joined. I was the first person to be hired whose main objective was business development. 

Why is your job as an evangelist important to Quantinuum? 

We can have the most amazing technology in the world, but if no one knows about it, then it doesn’t do anyone much good. There is a lot of misunderstanding and unfamiliarity that surrounds this industry currently, which is why my job of creating awareness is so important. 

I get to talk to university students and researchers and let them know we have software they can use for free to help them code better. I am very lucky to have an academic background in physics because when I speak at these universities, the professors sometimes let me take over the class for a day. I don’t think they would grant the same access to a salesperson. I love to talk about the cool things we have done and are doing with these students and share ways we can partner and collaborate both now and in the future.

We want to build our hiring pipeline with the smartest and most creative young minds available. Hiring is a top priority, and job candidates may not know there are such amazing job opportunities at Quantinuum and throughout this exciting industry.  

How has the industry changed in the last five years? 

When I started, there were 8–10 credible quantum computing startups, including us. We were all pretty small with just a few dozen employees at the time. 

Now, it seems like there’s a new company forming, a new investment, or a technical breakthrough in hardware or software every week. There are quantum information sciences degrees and programs in college now including quantum computing and closely related sciences. It’s dizzying to keep up with everything. 

Today, there are roughly 400 quantum companies, building quantum products all over the world. Companies are also increasing in size. Our company currently has 400 employees, but we’re hiring like crazy and anticipate adding 200 people in 2022. 

The U.S. government also is investing. During the last administration, they had a Quantum Initiative Act (QIA) where $1.2 billion was allocated for quantum funding. Other countries also are investing. China, for example, has spent at least $30 billion in quantum technology over the last few years.

We sat down with Brian Neyenhuis, Quantinuum’s director of commercial operations to ask him about the 20-qubit upgrade, some of the technical details, and how this launch paves the way for scaling trapped-ion quantum computers in the future.

What are some of the key upgrades made to the H1-1 machine?

The biggest, or maybe the most notable, is that we expanded the number of fully connected qubits from 12 to 20. That is a significant increase and the most qubits we’ve added to an existing machine. Last year, we added two fully connected qubits to the 10 qubits H1-1 already had. It was a major accomplishment at the time. Now, that seems easy compared to this upgrade because for us, it is not as simple as adding qubits.

To add eight more qubits and maintain all-to-all connectivity, we upgraded the optics that deliver the light used to control our qubits. Previously, we were only delivering the light needed to complete quantum gates to three different regions of the trap, which we call gate zones. Now we can address all five zones in our trap simultaneously. This enables us to complete more single-qubit or two-qubit gates in parallel, which means users can run complex algorithms without experiencing a slowdown.

How does this compare to previous hardware upgrades?

This one was significantly more involved than previous upgrades. Although we didn’t modify the trap at the heart of the computer or the vacuum chamber and cryostat that enclose it, we redesigned the entire optical delivery system. This was necessary so as not to deliver light to more regions of the trap, but also to improve stability.

Increasing the size and complexity of the machine without improving the stability would be a recipe for disaster. Because we were able to improve the stability, we were able to add more qubits without sacrificing performance or key features our users expect such as all-to-all connectivity, high single and two-qubit gate fidelities, and mid-circuit measurement.

Why is the increase in zones significant?

The gate zones are where all the interesting quantum stuff happens. More zones allow us to run more quantum operations in parallel, allowing for faster, more complex algorithms.

What's the connection between more zones and more qubits?

Having more gate zones allows us to use more qubits in an efficient way.

Because we can do all these operations in five different locations in parallel, it finally makes sense to put more qubits into the trap. We could have loaded more qubits into earlier versions of the system, but without additional gate zones, it doesn’t make a lot of sense. In fact, doing that would create a bottleneck with qubits waiting for their turn to do a two-qubit gate, which then slows down an algorithm. Now, we can do five quantum gates in parallel, which allows us to run more complex algorithms without sacrificing speed.

Twenty qubits are probably where this generation of traps ends. There is a possibility to add a handful more, but it feels like this is probably the most efficient number for these H1 Systems due to layout of the trap. But future generations, some of which are already trapping ions in the lab today, will use even more qubits and with the same or better efficiency.

What is the “ion dance”?

In the QCCD architecture, trapped ions are easy to move around. By applying the right set of voltages to the trap — a small, electrode-filled device that holds qubits in place — we can arbitrarily rearrange the chain of qubits so any qubit can pair with any other and perform a quantum gate. So, you can think of any algorithm as a set of steps where we shuffle all the qubits to pair them up for the next set of gates, move them into the gate zones, and then shuffle them again to set them up for the next set of gates. The ions “dance” across the trap moving from partner to partner to execute a quantum circuit.

Some circuits, like quantum volume circuits, are densely packed, meaning that every possible pair wants to do a gate at each step of the circuit. Other circuits are very loosely packed, meaning you can only do a few gates in parallel before moving on to the next slice because you need to reuse one of those qubits with a different partner.

Although this dance may sound complicated, it makes it very easy to program our quantum computer. A user sends us a time-ordered set of gates without having to think about the layout of the qubits, and our compiler figures out how to pair up the appropriate qubits to make it happen. You don't have to worry about which ones are next to each other because any pair of qubits is equal to all the others. And, at any step, we can completely rearrange this chain and put any two qubits next to each other.

It’s like a square dance where someone calls out directions to the dancers.

Anything else in the works for Quantinuum’s hardware this year?

We will continue to work with our customers to improve our system performance and their overall experience. One of the reasons we have a commercial system now is to allow our customers to program their algorithms on a real machine. They're dealing with all the constraints of real quantum hardware. They're pushing on their algorithms while we're pushing on the hardware, to get the fastest iterations.

As they learn new things about their algorithm, we learn what the most important things are to improve. And we work on those. We are learning a lot from our customers, and they are learning a lot by running on our hardware.

When it comes to completing the statistical tests and other steps necessary for calculating quantum volume, few people have as much as experience as Dr. Charlie Baldwin.

Baldwin, a lead physicist at Quantinuum, and his team have performed the tests numerous times on three different H-Series quantum computers, which have set six industry records for measured quantum volume since 2020.

Quantum volume is a benchmark developed by IBM in 2019 to measure the overall performance of a quantum computer regardless of the hardware technology. (Quantinuum builds trapped ion systems).

Baldwin’s experience with quantum volume prompted him to share what he’s learned and suggest ways to improve the benchmark in a peer-reviewed paper published this week in Quantum.

“We’ve learned a lot by running these tests and believe there are ways to make quantum volume an even stronger benchmark,” Baldwin said.

We sat down with Baldwin to discuss quantum volume, the paper, and the team’s findings.

How is quantum volume measured? What tests do you run?

Quantum volume is measured by running many randomly constructed circuits on a quantum computer and comparing the outputs to a classical simulation. The circuits are chosen to require random gates and random connectivity to not favor any one architecture. We follow the construction proposed by IBM to build the circuits.

What does quantum volume measure? Why is it important?

In some sense, quantum volume only measures your ability to run the specific set of random quantum volume circuits. That probably doesn’t sound very useful if you have some other application in mind for a quantum computer, but quantum volume is sensitive to many aspects that we believe are key to building more powerful devices.

Quantum computers are often built from the ground up. Different parts—for example, single- and two-qubit gates—have been developed independently over decades of academic research. When these parts are put together in a large quantum circuit, there’re often other errors that creep in and can degrade the overall performance. That’s what makes full-system tests like quantum volume so important; they’re sensitive to these errors.

Increasing quantum volume requires adding more qubits while simultaneously decreasing errors. Our quantum volume results demonstrate all the amazing progress Quantinuum has made at upgrading our trapped-ion systems to include more qubits and identifying and mitigating errors so that users can expect high-fidelity performance on many other algorithms.

You’ve been running quantum volume tests since 2020. What is your biggest takeaway?

I think there’re a couple of things I’ve learned. First, quantum volume isn’t an easy test to run on current machines. While it doesn’t necessarily require a lot of qubits, it does have fairly demanding error requirements. That’s also clear when comparing progress in quantum volume tests across different platforms, which researchers at Los Alamos National Lab did in a recent paper.

Second, I’m always impressed by the continuous and sustained performance progress that our hardware team achieves. And that the progress is actually measurable by using the quantum volume benchmark.

The hardware team has been able to push down many different error sources in the last year while also running customer jobs. This is proven by the quantum volume measurement. For example, H1-2 launched in Fall 2021 with QV=128. But since then, the team has implemented many performance upgrades, recently achieving QV=4096 in about 8 months while also running commercial jobs.

What are the key findings from your paper?

The paper is about four small findings that when put together, we believe, give a clearer view of the quantum volume test.

First, we explored how compiling the quantum volume circuits scales with qubit number and, also proposed using arbitrary angle gates to improve performance—an optimization that many companies are currently exploring.

Second, we studied how quantum volume circuits behave without errors to better relate circuit results to ideal performance.

Third, we ran many numerical simulations to see how the quantum volume test behaved with errors and constructed a method to efficiently estimate performance in larger future systems.

Finally, and I think most importantly, we explored what it takes to meet the quantum volume threshold and what passing it implies about the ability of the quantum computer, especially compared to the requirements for quantum error correction.

What does it take to “pass” the quantum volume threshold?

Passing the threshold for quantum volume is defined by the results of a statistical test on the output of the circuits called the heavy output test. The result of the heavy output test—called the heavy output probability or HOP—must have an uncertainty bar that clears a threshold (2/3).

Originally, IBM constructed a method to estimate that uncertainty based on some assumptions about the distribution and number of samples. They acknowledged that this construction was likely too conservative, meaning it made much larger uncertainty estimates than necessary.

We were able to verify this with simulations and proposed a different method that constructed much tighter uncertainty estimates. We’ve verified the method with numerical simulations. The method allows us to run the test with many fewer circuits while still having the same confidence in the returned estimate.

How do you think the quantum volume test can be improved?

Quantum volume has been criticized for a variety of reasons, but I think there’s still a lot to like about the test. Unlike some other full-system tests, quantum volume has a well-defined procedure, requires challenging circuits, and sets reasonable fidelity requirements.

However, it still has some room for improvement. As machines start to scale up, runtime will become an important dimension to probe. IBM has proposed a metric for measuring run time of quantum volume tests (CLOPS). We also agree that the duration of the computation is important but that there should also be tests that balance run time with fidelity, sometimes called ‘time-to-solution.”

Another aspect that could be improved is filling the gap between when quantum volume is no longer feasible to run—at around 30 qubits—and larger machines. There’s recent work in this area that will be interesting to compare to quantum volume tests.

You presented these findings to IBM researchers who first proposed the benchmark. How was that experience?

It was great to talk to the experts at IBM. They have so much knowledge and experience on running and testing quantum computers. I’ve learned a lot from their previous work and publications.

There is a lot of debate about quantum volume and how long it will be a useful benchmark. What are your thoughts?

The current iteration of quantum volume definitely has an expiration date. It’s limited by our ability to classically simulate the system, so being unable to run quantum volume actually is a goal for quantum computing development. Similarly, quantum volume is a good measuring stick for early development.

Building a large-scale quantum computer is an incredibly challenging task. Like any large project, you break the task up into milestones that you can reach in a reasonable amount of time.

It's like if you want to run a marathon. You wouldn’t start your training by trying to run a marathon on Day 1. You’d build up the distance you run every day at a steady pace. The quantum volume test has been setting our pace of development to steadily reach our goal of building ever higher performing devices.

Quantinuum today honored researchers from the National Institute of Standards and Technology (NIST) for their technical achievements and contributions to the field of quantum computing.

In a ceremony at the company’s U.S. headquarters in Broomfield, President and Chief Operating Officer Tony Uttley recognized the decades of innovative research by NIST’s Ion Storage Group and the role it has played in the development of Quantinuum’s H-series hardware technology, which recently set an industry record for performance.

“It’s impossible to overstate the impact of the NIST Ion Storage Group and its research,” Uttley said. “Quantum computing has advanced to where it is today in large part because of this group and its commitment to making its work available. Their research forms the basis for the trapped ion quantum computing technologies being developed by Quantinuum and others. It is truly a technology transfer success story for the U.S. government.”

NIST’s Colorado-based ion trap group was formed in the late 1970s not long after Dr. David Wineland, demonstrated that by using lasers, it was possible to cool ions to low enough temperatures that they could be manipulated and controlled while trapped in electromagnetic fields.

This discovery and the team’s subsequent research led to the development of some of the world’s most precise atomic clocks, a technology that helps enable Global Positioning Systems (GPS) satellites.

In the 1990s, the NIST group expanded its focus to quantum information processing and quantum computing. In 1995, the NIST team successfully executed the world’s first entangling two-qubit quantum gate, an operation that is key to quantum computing.

In 2000, the group demonstrated for the first time the more robust Mølmer-Sørensen gate, entangling four ion qubits. The Mølmer-Sørensen gate is at the heart of almost all ion-trap quantum computing gates today.

In 2002, the team published an article in Nature outlining the concept of the Quantum Charged Coupled Device (QCCD) architecture for a trapped ion-based quantum computer. (Quantinuum uses this QCCD architecture in its H-Series hardware, Powered by Honeywell.)

These advancements and others led to Wineland sharing the 2012 Nobel Prize for Physics with Serge Haroche for "ground-breaking experimental methods that enable measuring and manipulation of individual quantum systems.

The NIST team continues to advance trapped ion technologies. Quantinuum recently signed an agreement with NIST to collaborate on some trap design elements.

Uttley said Quantinuum’s relationship with NIST is critical to the company’s success and its ongoing efforts to build the highest performing quantum computers in the world.

“The NIST team has a deep expertise in ion trap design, which will continue to help us on the technical side,” Uttley said. “The agency also has trained a great number of students and researchers over the years to become leading experts in the field and helped bolster the current and future quantum workforce.”

“Technology transfer is an important way that NIST achieves its mission of promoting U.S. innovation and industrial competitiveness,” said Director of NIST’s Physical Measurement Laboratory Jim Kushmerick. “We are always excited to see our research applied to develop commercial products, particularly those with great potential such as quantum computing.”