Paul Lipman is Chief Revenue Officer at Infleqtion, leading global growth efforts at the cutting edge of quantum technology.
In the modern world, we take for granted the transistors that enable everything from cell phones to supercomputers. However, in the early days of computing, the outcome was far from clear, as computing technology evolved through mechanical systems, electromechanical relays, vacuum tubes and other techniques before transistors ultimately became the standard.
In the 1950s, one of the world’s largest computers was the AN/FSQ-7, a Cold War air defense system that weighed 250 tons, consisted of 49,000 vacuum tubes and drew 3 MW of power. Vacuum tube systems were ultimately eclipsed by transistors, which proved cheaper, faster, smaller, more reliable and manufacturable at scale. Today’s flagship cell phones each contain tens of billions of them.
Quantum computing today feels somewhat akin to classical computing before the advent of the transistor. Popular approaches include superconducting circuits, trapped ions, neutral atoms and photonics. We have not yet converged on a dominant platform in large part because qubits (quantum bits) are far harder to build and control than classical bits. Qubits must be precisely controllable, protected from noise that can decohere their delicate quantum state, entangleable with one another and scalable into practical, error-corrected systems. The platform that ultimately leads the quantum computing industry will not do so on performance alone—it will also need to address manufacturability, energy efficiency, facility footprint and commercial relevance.
Neutral atom quantum computers trap individual atoms in arrays of laser light and use additional lasers to control their quantum states. Because the qubits are individual atoms, we effectively get them free from nature, and they are intrinsically identical in a way that fabricated qubits are not. Atoms also offer key advantages in scalability, connectivity and dramatically lower infrastructure and power requirements. Small and electrically neutral, atomic qubits can be packed very closely together in dense two-dimensional arrays. Our company set the industry record at 1,600, and academic groups have now demonstrated over 6,000. Unlike superconducting or photonic approaches that will require football field-sized facilities to achieve practical scale systems, a million neutral atom qubits could be contained within a quantum core only inches on a side. The surrounding lasers, optics and electronics necessary for controlling the atoms are all highly amenable to miniaturization.
Qubits must be kept extremely cold to preserve their fragile quantum states from environmental noise. To achieve temperatures close to absolute zero, superconducting and photonic systems require large, power-hungry, cryogenic infrastructure—utility-scale systems using these modalities will require tens of megawatts. In contrast, neutral atom quantum computers utilize laser cooling techniques that require orders of magnitude less power and small form factor systems.
Thomas Watson is said to have remarked in 1943, “I think there is a world market for maybe five computers.” While this quote is almost certainly apocryphal, the lesson is real: Platform shifts always look modest before they scale. As we think about broad-scale deployment of quantum computers, not just in national labs and institutions, but in cloud infrastructure and data centers, the size, power and environmental advantages of neutral atoms will make a pivotal difference.
Neutral atoms also offer key architectural benefits. In contrast to architectures with fixed connectivity, atomic qubits can be both entangled in place and dynamically moved to suit the needs of computation and error-correction. That means connectivity is not simply a static property of the system but rather software-defined and dynamically controllable, reducing the need for inefficient codes that must contort themselves around fixed system architectures.
Atoms are exquisitely sensitive to specific wavelengths of light, which enables precise control of their quantum states. This enables us to use atoms of different species within a single quantum computer, where one species performs computation and another assists with measurement, reset or error-correction workflows. This makes mid-circuit measurement, a critical element of fault-tolerant quantum computing, faster and more efficient.
Of course, no computational modality is without its shortcomings. For example, neutral atoms are often described as having a slower “clock speed” than superconducting qubits. However, atomic qubits’ scalability and connectivity enable parallelism and a richer array of efficient error correction codes that substantially close the gap in terms of overall wall time to complete a given computation.
Neutral atoms also have several less obvious, but no less important, advantages.
Neutral atoms appear unusually capital efficient relative to many competing approaches. The hardest engineering challenges are in optics, photonics, beam control and vacuum systems—areas that can be improved, reused and manufactured across large adjacent markets. That makes for a significantly more efficient model than one built around ever-larger bespoke infrastructure. Neutral atoms therefore offer not just a plausible path to strong computational performance but a more compelling path to commercial relevance and scale.
Many of the technologies that enable neutral atom quantum computing also enable quantum sensing, including atomic clocks, and RF, inertial and gravitational sensors. This provides tremendous architectural and business leverage as improvements in lasers, beam steering, vacuum systems, photonics and field control can benefit multiple quantum product lines at once.
Looking further ahead, a common neutral atom platform may ultimately allow sensing qubits and computational qubits to coexist within the same system. If so, it could become possible in some applications to preserve sensed data in its quantum format, for processing by a quantum algorithm. That may open entirely new system architectures and computational advantages, some of which have been explored in academic work.
This does not mean the race is over. Quantum computing has not yet had its transistor moment. As an industry, we are still in the age of the AN/FSQ-7. But when that moment comes, the winner will not simply be the modality with the loudest headlines or the largest funding rounds. It will be the one that best combines performance, manufacturability, upgradeability, energy efficiency, connectivity and deployability into a true commercial platform.
My money is on neutral atoms.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
