Trapped-ion quantum computing uses individual atoms that have been ionized (had an electron removed) and confined in electromagnetic traps, typically Paul traps that use oscillating radio-frequency fields to create a stable potential well. The qubit states are encoded in two energy levels of the ion's electronic structure — either hyperfine ground states (as in ytterbium-171, used by Quantinuum and IonQ) or an optical transition (as in calcium-40 or barium-138). Because every ion of the same species is identical, trapped-ion qubits are naturally uniform with no fabrication variability.
Trapped ions hold the records for single-qubit and two-qubit gate fidelities: above 99.9999% for single-qubit gates and above 99.9% for two-qubit gates (Quantinuum). Coherence times reach seconds to minutes for hyperfine qubits, orders of magnitude longer than superconducting alternatives. Two-qubit gates are implemented via the Coulomb interaction between ions in the same trap, using laser or microwave pulses to create entanglement through the ions' shared motional modes (Molmer-Sorensen or light-shift gates).
The main scaling challenge for trapped-ion systems is speed and connectivity as system size grows. Gate times are microseconds (1,000x slower than superconducting gates), and ions in a single trap interact through shared motional modes that become crowded as ion count increases beyond roughly 30-50 per trap zone. Quantinuum and IonQ address this through ion transport architectures (QCCD) that shuttle ions between zones, enabling all-to-all connectivity at the cost of transport time overhead. Trapped-ion systems are strong candidates for early fault-tolerant quantum computing due to their exceptional gate quality.