Neutral atom quantum computing

Imagine a computer that does not process information by flipping switches between zero and one, but instead holds both possibilities at once, like a coin spinning in mid-air that is simultaneously heads and tails. This is the promise of the quantum computer, a machine that leverages the strange rules of nature to solve problems that would take ordinary machines thousands of years to crack. To build such a machine, scientists must create tiny, stable units of information called qubits that can be controlled with extreme precision, linked together to form complex networks, and protected from the slightest disturbance that would cause them to collapse. For decades, researchers have explored many ways to build these units, from superconducting circuits to trapped ions, but one approach has recently surged forward with remarkable speed: using individual atoms of gas, cooled to temperatures near absolute zero and held in place by beams of light, to serve as the building blocks of a quantum processor.

In a comprehensive review published in 2026, Mark Saffman of the University of Wisconsin-Madison and Infleqtion details the rapid evolution of this neutral atom approach. The paper traces the journey from theoretical proposals made over twenty-five years ago to the current reality of massive arrays containing thousands of atoms, each acting as a qubit. The core of this technology relies on trapping atoms in a vacuum using focused laser beams, creating a grid of light that holds the atoms in place without touching them. Once trapped, these atoms can be manipulated with lasers to store information, perform calculations, and link with one another. The review highlights that while early experiments demonstrated the basic ability to control single atoms, recent years have seen a dramatic leap in scale and performance, with researchers now successfully running quantum algorithms and creating logical qubits that can correct their own errors, a critical step toward building a machine that can operate reliably for long periods.

The choice of which atom to use is a fundamental decision in this field, and the paper explains that scientists have settled on a few specific types, primarily rubidium and cesium, which are heavy alkali metals, as well as strontium and ytterbium, which belong to a different family of elements. These atoms are chosen because their internal energy levels can be precisely controlled and because they respond well to laser cooling. The process begins by slowing the atoms down until they are nearly motionless, a state where they can be captured by the gentle pressure of laser light. The researchers use optical tweezers, which are essentially tightly focused beams of light that act like invisible fingers to hold individual atoms in a specific pattern. In the most advanced setups, these tweezers can arrange thousands of atoms into a perfect grid, with some recent demonstrations showing arrays with over ten thousand sites, and even larger systems approaching one hundred thousand sites using specialized optical components.

Once the atoms are trapped, the next challenge is to encode information into them. Scientists do this by selecting two specific energy states within the atom to represent the zero and one of a qubit. For some atoms, these states are chosen because they are naturally resistant to magnetic noise, allowing the information to remain stable for many seconds. For others, the information is stored in a transition between the ground state and a long-lived excited state, which requires a special type of trapping light that does not disturb the delicate balance between the two states. The paper notes that while rubidium and cesium atoms have been used for a long time, the newer strontium and ytterbium atoms offer significantly longer periods of stability, keeping their quantum information intact for several seconds without the need for complex correction techniques. This stability is crucial because it gives the computer enough time to perform many calculations before the information fades away.

To perform a calculation, the atoms must be able to talk to one another. The paper describes two main ways this happens. The first involves moving the atoms close together so they collide, but this method is difficult to control precisely. The second, and far more successful method, involves exciting the atoms to a high-energy state known as a Rydberg state. In this state, the atom becomes enormous, with its outer electron orbiting far from the nucleus, causing it to interact strongly with neighboring atoms. This interaction creates a "blockade" effect: if one atom is excited, it prevents its neighbors from being excited at the same time. By carefully timing laser pulses, researchers can use this blockade to create a link between two atoms, allowing them to perform a logical operation that entangles their states. Recent experiments have shown that these entangling gates can be performed with extremely high accuracy, reaching fidelities above ninety-nine percent, which is close to the level needed for practical, error-corrected computing.

A major hurdle for any quantum computer is that errors are inevitable. The paper explains that to overcome this, scientists are developing systems that can detect and fix mistakes as they happen. This is done by grouping physical atoms together to form a single "logical" qubit that is more robust than any individual atom. If one atom in the group loses its information or disappears, the system can detect the loss and replace it with a fresh atom from a reserve, or correct the error using the remaining atoms in the group. The review highlights that neutral atom systems have a unique advantage here: because the atoms are held by light, they can be moved around easily. This allows researchers to physically transport atoms to different parts of the machine for cooling or replacement without stopping the computation. Recent demonstrations have shown logical qubits with error rates that improve as the system grows larger, a sign that the technology is moving toward the threshold required for fault-tolerant operation.

The paper concludes by looking at the path forward, acknowledging that while the progress has been spectacular, significant engineering challenges remain. Scaling these systems to the hundreds of thousands of qubits needed for truly useful applications will require new optical technologies to manage the heat generated by the lasers and to control the movement of atoms with greater precision. The author suggests that the future may lie in modular designs, where several smaller arrays are linked together to form a larger machine, or in the development of new types of error-correcting codes that take advantage of the long-range connections possible with these atoms. Despite the difficulties, the review paints a picture of a field that has moved from theoretical possibility to experimental reality, with neutral atoms now standing as one of the most promising candidates for building the quantum computers of the future.