Part One · Quantum computing
Quantum hardware compared by qubit type
Hardware landscape
No single qubit modality has won out. Superconducting circuits, neutral atoms, trapped ions, and photonics are all being pursued in parallel, each trading off fidelity, connectivity, scalability, and operating conditions differently.
| Modality | Example system | Reported figures |
|---|---|---|
| Superconducting | Google Willow | 99.88% two-qubit gate fidelity; completed a benchmark random-circuit-sampling task in under five minutes, which Google states would take one of today's fastest classical supercomputers dramatically longer to reproduce |
| Superconducting | IBM (Kookaburra roadmap) | Targeting a 4,158-qubit system built from a multi-chip processor architecture |
| Superconducting | Rigetti Cepheus-1 | 108 qubits, 99.5% fidelity, general availability targeted for early 2026 |
| Neutral atom | QuEra | Targeting 100 logical qubits atop roughly 10,000 physical qubits by 2026–2027 |
| Photonic | PsiQuantum Omega | 99.98% single-qubit state preparation fidelity; 99.22% two-qubit fusion fidelity |
| Photonic | Xanadu Borealis / Aurora | Borealis: 216 squeezed-state qubits for Gaussian boson sampling demonstrations. Aurora: 12-qubit universal, gate-based architecture aimed at fault tolerance |