Google Quantum AI · The Techno Optimist

Analyst view

Willow (105 qubits, Dec 2024): logical errors halved with each step up in code size.
Quantum Echoes (Oct 2025) ran a physics measurement about 13,000x faster than the Frontier supercomputer.
On a 6-step roadmap. Steps 1 and 2 are done. Now working on a long-lived logical qubit.

As reported

Qubits

Transmon qubits and couplers sit on a square grid, which suits the surface code's neighbour-to-neighbour checks.

QEC

Helper qubits keep checking their data-qubit neighbours. Changes in the checks reveal errors, which a decoder then fixes.

Quality first

Willow's gains came from better materials that tripled how long qubits hold their state, to about 68 µs.

Applications

Google focuses on computations whose answers can be verified, for example against lab scans of molecules (NMR).

Programs

Google Quantum AI, one program

Updated 22 Oct 2025

Research · Proto · Pilot · Deploy · Scaling

5 items

Oct 2025
paperQuantum Echoes: out-of-time-order correlator measurement on Willow (opens nature.com)Willow and error-correction roadmap
A verifiable beyond-classical computation with a proposed route to molecular structure measurements.
May 2025
paperHow to factor 2048 bit RSA integers with less than a million noisy qubits (opens arxiv.org)
Craig Gidney cut the estimated resources to break RSA-2048 by about 20x, informing post-quantum migration timelines.
Dec 2024
paperQuantum error correction below the surface code threshold (opens nature.com)Willow and error-correction roadmap
Willow's largest logical qubit cut errors by over 2x at each step up in code size.
Feb 2023
paperSuppressing quantum errors by scaling a surface code logical qubit (opens nature.com)Willow and error-correction roadmap
A bigger code slightly beat a smaller one, the first sign that errors drop with scale.
Oct 2019
paperQuantum supremacy using a programmable superconducting processor (opens nature.com)Willow and error-correction roadmap
53-qubit Sycamore sampled random circuits in 200 seconds, a task estimated to take classical supercomputers far longer.

Open · Active · Partial

Correlated errorsPartially solved
Cosmic rays and qubits leaking out of their states cause bursts of errors across many qubits, which the surface code handles poorly.
From one logical qubit to manyOpen
A useful computer needs hundreds to thousands of logical qubits: about a million physical qubits by current estimates.
Useful applicationsActive research
Beyond-classical results so far are physics experiments. Finding useful commercial problems for modest machines is still open.

Company-wide

2029Large error-corrected quantum computer (end-of-decade target)Willow and error-correction roadmapby ~39 mo

Superconducting · 3

3 partners

Partner
NASA Ames (opens nasa.gov)
Co-founded the Quantum Artificial Intelligence Lab in 2013; co-author on the 2019 result.
Partner
QuEra
Google invested in QuEra in 2025.
Partner
UC Santa Barbara (opens ucsb.edu)
Hardware team originated in John Martinis's group.

Key people