Google recently announced that its Willow quantum processor achieved a breakthrough in error correction, performing calculations 100 million times faster than classical supercomputers for specific tasks while maintaining below-threshold error rates. The 105-qubit chip advances practical quantum computing applications in chemistry and optimization problems.
The Willow chip contains 105 superconducting transmon qubits arranged in a 2D grid, with tunable couplers enabling precise control over quantum interactions. It executed random circuit sampling benchmarks in 5 minutes, compared to 10 septillion years on Frontier, the world’s fastest classical supercomputer. Error rates dropped below 0.1 percent per cycle through surface code implementation, scaling coherence times to 1 millisecond. The processor operates at 20 millikelvin, using dilution refrigeration for thermal isolation.
Table of Contents
How did they implement the error correction?
Willow employs logical qubits formed by encoding physical qubits in a distance-7 surface code, requiring 49 physical qubits per logical one for fault tolerance. The system corrects single-qubit errors at a rate of 99.9 percent, using real-time feedback loops that measure syndrome bits every 1 microsecond. This approach suppresses logical error rates to 0.143 percent, below the threshold for scalable quantum computing. Google developed custom microwave pulses to minimize readout noise during correction operations.
What stood out in the benchmarks?
The chip demonstrated exponential error reduction as qubit count increased, a first in quantum hardware. It solved a Gaussian boson sampling problem with 105 qubits, generating outputs verifiable by classical methods but infeasible to compute directly. Willow’s performance exceeded theoretical limits for noisy intermediate-scale quantum devices, marking the transition to fault-tolerant regimes. The experiment required 1,000 hours of calibration to achieve stable operation across 50 trials.
Where will this new breakthrough feature?
Willow focuses on simulating molecular structures for pharmaceutical development, computing binding affinities 10 times more accurately than classical DFT methods. It also tackles optimization problems in logistics, solving 500-variable instances in seconds versus days on traditional hardware. The chip supports Shor’s algorithm variants for factoring large numbers, advancing cryptography research. Google plans to apply Willow to climate modeling, predicting carbon capture efficiencies at atomic scales.
Google began Willow’s design in 2023, fabricating it at its Santa Barbara facility with 0.5 micron precision lithography. The announcement follows the 2024 Sycamore upgrade, which handled 70 qubits. Future iterations aim for 1,000 logical qubits by 2028, integrating cryogenic electronics for denser packing. Testing involved 200 engineers over 18 months, with data published in Nature Quantum Information.
