Quantum Hardware Database
Google Willow
Google Willow succeeded Sycamore as Google's flagship superconducting processor, notably demonstrating that logical error rates can decrease as more physical qubits are added to a surface code — a long-sought milestone suggesting quantum error correction can scale as theory predicts.
Qubit Count
105
Qubit Type
Superconducting (transmon qubits)
Architecture
2D grid with nearest-neighbor connectivity
Released
2024
Gate Error (2Q)
~0.1–0.3% per two-qubit gate
Coherence Time
T₁ approaching 100 µs, improved over Sycamore generation
Operating Temp
~15 millikelvin
Cloud Access
Not publicly available
Company
Google (Google Quantum AI)
Connectivity
2D rectangular grid — each qubit connects to 4 nearest neighbors
Key Features
- Demonstrated 'below threshold' error correction scaling, a key fault-tolerance milestone
- Significant improvements in coherence time over the Sycamore generation
- Used to run improved random circuit sampling benchmarks
- Built with refined fabrication techniques to reduce defect-related errors
Limitations
- Not available for public cloud access
- Still well below the qubit counts needed for large-scale fault-tolerant algorithms
Technical Notes
Willow's most significant result was showing that, as the surface code's distance is increased (more physical qubits protecting each logical qubit), the logical error rate decreases exponentially — confirming a foundational prediction of quantum error correction theory at meaningful scale for the first time.