← All quantum technologies

computing platform profile

Neutral-atom quantum computing

Neutral-atom computers arrange individual atoms into programmable arrays with tightly focused light. Their geometry can be reconfigured for digital gates or analog quantum simulation.

  • optical tweezers
  • Rydberg states
  • quantum arrays

Technical deep dive

How it works

Optical tweezers trap and move atoms into a selected pattern. Lasers address internal atomic levels, and temporary excitation to large Rydberg states produces strong interactions between nearby atoms for entangling operations. The platform must maintain atom loading, laser stability, calibration, readout fidelity, and fast repeated cycles while scaling both qubit number and circuit depth.

Assembling an atomic register

Far-detuned focused laser beams create optical tweezers whose intensity gradient traps individual atoms. Initial loading is probabilistic, so fluorescence imaging identifies occupied sites and moving tweezers rearrange atoms into a defect-free target geometry. Additional reservoirs can replace lost atoms between circuit stages.

Qubits use two long-lived hyperfine, clock, or metastable internal states. The atoms are intrinsically identical, but their local light shifts, temperature, motion, magnetic field, and addressing intensity are not. Calibration must keep transition frequencies and Rabi rates uniform across a large field of view.

Rydberg blockade and gates

Lasers temporarily couple selected atoms to highly excited Rydberg states. Their large dipole interactions shift the Rydberg level of nearby atoms, preventing simultaneous excitation inside a blockade radius. Controlled pulse sequences turn that conditional excitation into an entangling gate; analog simulators instead map Rydberg excitation and interactions directly onto a many-body Hamiltonian.

Gate error comes from finite Rydberg lifetime, spontaneous emission through intermediate states, Doppler motion, laser phase and amplitude noise, imperfect blockade, crosstalk, leakage, and atom loss. Faster gates reduce decay but demand higher laser power and can increase off-resonant error. Geometry is reconfigurable, but movement adds time and heating.

Scale versus computational depth

Large atom arrays are compelling for quantum simulation and offer spare atoms for error correction. Yet array size measures trapped sites, not the number that can receive a complete high-fidelity digital circuit. A machine needs parallel addressing, non-destructive or replaceable readout, repeated syndrome extraction, low-latency decoding, and a cycle time short relative to loss and decoherence.

Imaging commonly ejects atoms or destroys the encoded state, so mid-circuit measurement and reset require careful architecture. Logical performance will depend on correlated laser errors, spatially varying loss, decoder behavior, and the ability to reload without disturbing data qubits.

Reason to use it

What distinguishes the approach

Atoms of the same isotope are intrinsically uniform, while optical tweezers can assemble large, reconfigurable geometries without fabricating a wire to every site.

Unresolved engineering

Main scaling problem

Combining large arrays with low error and long, repeatable circuits

Monthly evidence layer

Research watch

Named programs, dated developments, and primary or official sources. Company claims are identified separately from independent or peer-reviewed evidence.

Last reviewed 6 September 2026

Research groups & programs

Current developments

  1. 2025

    Coherent atom arrays crossed the several-thousand-site scale

    The achievement expands simulation and redundancy resources while making uniform trapping, imaging, and control a system-scale problem.

  2. 2025

    Entangling fidelity rose while exposing the role of loss filtering

    Reporting raw and postselected values separately gives a clearer view of what error correction would actually receive.

Papers & official resources