Neutral-atom qubits

Neutral-atom qubits encode quantum information in the internal states of individual neutral atoms, such as rubidium or cesium, held in place by tightly focused laser beams called optical tweezers. Arranged into two- or three-dimensional arrays, the atoms are entangled by briefly exciting them to high-energy Rydberg states, giving a modality prized for large qubit numbers and reconfigurable geometry.

#How they work

Laser cooling loads atoms into an array of optical tweezers, each trap holding a single atom. Two long-lived atomic levels form the Qubit states, and global or focused laser and microwave fields drive single-qubit rotations. Two-qubit gates exploit the Rydberg blockade: when an atom is promoted to a Rydberg state, its enormous electric dipole moment shifts the energy levels of nearby atoms enough to prevent them from being excited simultaneously, which produces controllable entanglement between atoms within a blockade radius (Saffman, Walker, Mølmer 2010). Because tweezers can be moved with acousto-optic deflectors, atoms can be physically rearranged mid-computation, giving flexible and even dynamic connectivity.

#Strengths

• Arrays of hundreds to more than a thousand atoms have been assembled, and the atoms are naturally identical, similar to Trapped-ion qubits.
• The layout is reconfigurable: qubits can be shuttled to bring distant atoms together, reducing the routing overhead of fixed lattices.
• The same platform is a powerful analog quantum simulator, first shown at scale on a 51-atom system studying many-body dynamics (Bernien et al. 2017).

#Limitations

• Gate fidelities, while improving quickly, have historically trailed the best trapped-ion results, and Rydberg excitation is sensitive to laser noise and atomic motion.
• Atoms are occasionally lost from traps and must be detected and reloaded, adding overhead.
• Measurement and reloading cycles can be slower than gate-model solid-state systems.

The finite lifetime of the Rydberg state and imperfect addressing lasers set practical limits on how many high-fidelity gates a circuit can contain. Preparing and imaging large arrays also takes time, so the effective clock speed of a full computation is set as much by loading, cooling, and readout as by the gates themselves. These are engineering constraints rather than fundamental barriers, and they have improved steadily as control hardware matures.

#Progress toward logical qubits

A 2023 experiment used reconfigurable atom arrays to operate dozens of logical qubits and run small error-corrected circuits, a notable demonstration of Quantum error correction on this platform (Bluvstein et al. 2023). It illustrated how mid-circuit atom movement can implement the transversal operations that error-correcting codes need.

#Who is building it

Pasqal builds neutral-atom processors and quantum simulators based on Rydberg arrays and offers cloud and on-premises access (Pasqal). Other groups and companies pursue the same modality, and published results are best read by separating demonstrated experiments from company roadmaps.

#Status

As of early 2026 neutral-atom systems combine some of the largest qubit registers with rapidly rising fidelity and early logical-qubit demonstrations, making them a serious contender among modalities. Like all current hardware they remain in the NISQ era with no demonstrated advantage on a cryptographically meaningful problem, so they neither accelerate nor settle the Q-Day timeline. Contrived sampling tasks aside, useful Quantum advantage has not been shown on this or any other platform.