From Neutral Atoms to Quantum Utility: Infleqtion's Vision for the Future of Quantum Computing

Marco Palumbo , Infleqtion UK 's Director of Business Development, interviews Thomas Noel , Infleqtion 's VP of Quantum Computing, about Infleqtion 's Quantum Computing platform and plans for the future. Read on to learn more about the unique features of the platform and the potential applications, plus the distinction between analog and gate-model quantum computing.

Marco Palumbo : Hi, Tom. Could you tell us a bit about yourself? What do you do at Infleqtion ?

Thomas Noel : Hi, Marco. And welcome to Infleqtion . Off the clock, I'm an avid rock climber, an increasingly avid mountaineer, and a mixology enthusiast. But that's probably not what you meant. Professionally, I'm a physicist and currently Infleqtion ’s VP for Quantum Computing. Deep in my past, at the University of Washington, I worked on experiments employing trapped barium ions for quantum communication. That ultimately led me to Infleqtion , where I started extending our world-class neutral atom vacuum cell technology for use with ion traps. More recently, I've been leading the team building Infleqtion's neutral atom quantum computing systems.

Marco Palumbo : Tell us more about the work Infleqtion is doing in Quantum Computing. What's different from any other platform out there??

Thomas Noel : Infleqtion ’s quantum computer is a neutral atom quantum computer, which means we encode quantum information in electronic states of individual atoms, in our case – Caesium atoms. So, each atom corresponds to one "quantum bit" or qubit. This neutral atom qubit technology is a main feature that distinguishes our approach. Many other companies out there working in this space are working with other types of qubits (trapped ions, superconducting circuits, photons, etc.). We believe neutral atoms as qubits are promising for achieving scale and, ultimately, fault-tolerant operations, so let me say a little bit about that.?

We love neutral atoms as qubits because atoms of a given species are all naturally identical (avoiding materials challenges associated with engineered qubit types) and because appropriately chosen atomic states can be incredibly insensitive to environmental noise, making quantum information stored in them intrinsically long-lived.

Finally, and this one is a bit of a fine technical point, we like that the "Rydberg interaction" mechanisms that enable neutral atom qubit "entangling" operations are independent of how many atoms are around. This feature largely de-couples the challenges associated with scaling up qubit count from the challenges associated with improving the performance ("fidelity") of multi-qubit entanglement. When the team decided to focus on scaling up, in just three months, we were able to increase the qubit count by a factor of six without any reduction in gate fidelities or shot rate. I don't want to make it sound too easy, though!

Neutral atoms provide us with plenty of high-quality qubits, but we trade that for optical and laser engineering challenges. The quantum operations that make our quantum computers compute entail precise atomic state manipulation by carefully controlled lasers. Infleqtion has deep expertise in photonic technology and will be driving some significant near-term innovations in this critical area for quantum computing.

Within quantum computing, broadly speaking, there are two models for generating interesting quantum states that encode the answers to questions that are not accessible to classical computers. There is the "analog" approach on the one hand and the "digital" or "gate-model" approach on the other hand. Infleqtion is focused on the gate-model approach while keeping an option on the analog mode also.

Marco Palumbo : Analog neutral quantum computing, gate-model neutral quantum computing. What is the difference? Which one is best at what?

Thomas Noel : Both analog and gate-model approaches aim to do the same thing: generate interesting quantum states that, when measured, tell us something useful. The way that each approach generates those quantum states is what is different.

An analog quantum computer relies on the similarity between its underlying physical interactions and those that naturally produce a given desired type of quantum state. For example, a city planner might want to know the best distribution of traffic management devices to facilitate efficient travel within a neighborhood. This kind of question can be boiled down into what mathematicians refer to as a "combinatorial optimization problem".

Suppose the constraints and interactions in the optimization problem are very similar to the underlying physics of an analog quantum computer. In that case, that analog quantum computer may be able to efficiently find solutions. A gate-model quantum computer can also solve these kinds of problems by "discretizing" the generation of the desired quantum state into a sequence of gate operations. This discretization may be less efficient in some cases, leading to deep and difficult-to-execute "quantum circuits."

But even seemingly small tweaks to the combinatorial problem may make the underlying physics of the analog quantum computer inappropriate for finding solutions. And for other quantum computing use cases, the analog approach just doesn't map well. In such cases, a gate-model quantum computer, which is "universal," will still be able to attack the problems.

This is the basic trade-off in the near term: analog quantum computers will be able to efficiently target a niche use case while gate-model processors have the potential to be transformative across a broad range of applications.

Furthermore, the concept of quantum error correction looks critical to the long-term success of quantum computing and is most well-established for gate-model quantum computers. So, in the long term, we believe that gate-model quantum computing will be the predominant approach. Prioritizing universality, market reach, and error correction has led us to focus on the gate-model approach at Infleqtion , even though our core technology can equally well be used for analog quantum computing in principle.

Marco Palumbo : Let's talk about applications. And timescale. What is Infleqtion's quantum computer good for right now? What are people using it for? Where do you see the major area of applications in the future, and how long do we need to wait to see its full potential being deployed?

Thomas Noel : Infleqtion ’s quantum computer is accessible over the cloud today, and we are working with select partners to run their applications already. These early applications are in the realm of materials simulation and chemistry. Of course, like all quantum computers today, these applications are being investigated at a "small scale" where the quantum computer is not yet outperforming alternative classical approaches to finding the same answers. Still, we are learning a huge amount from these interactions. They help us to improve our quantum computer and our user interfaces. Plus, they help focus our development roadmap on the capabilities that will be most enabling for applications in the years to come.

You ask how long it will be before the full potential of quantum computing is deployed, but I don't see it that way. Just as classical computers remain under development, quantum computers will continue to develop in the near future, their computational power continuing to grow, and new application areas continuing to emerge as we find new ways for quantum processors to complement classical processors. But there is an important day coming when quantum computers will outperform classical computers on tasks of practical importance – a feat some have referred to as "quantum utility" or "quantum practicality." As they say, it is difficult to make predictions, especially about the future. Still, this day seems to me to no longer be a decade out, but a matter of a few years. We believe the neutral atom platform is the fastest path to getting there, and we're on our way!

Marco Palumbo : Thanks a lot, Tom. It has been a pleasure talking to you. I do feel I understand the subject more now.

Learn more: A simple, passive design for large optical trap arrays for single atoms by Mark Saffman , P. Huft, Y. Song, T. M. Graham, K. Jooya, S. Deshpande, C. Fang, M. Kats.?