Skip to content
ContentLora

    Tip: press / anywhere to search.

    product

    Google Willow

    Also known as Willow chip, Google Willow processor

    Willow is a 105-qubit superconducting quantum processor that Google announced in December 2024.[1] It was the first chip to show error correction getting better as the code grew, and in 2025 it ran Quantum Echoes, which Google calls the first verifiable quantum advantage.[2][3][4]

    Editor reviewedUpdated Quantum computingComputingPhysics
    Key facts

    Overview

    Willow is Google Quantum AI’s 105-qubit superconducting processor, announced on 9 December 2024.[1] It was built in Google’s dedicated quantum fabrication facility in Santa Barbara, California, and its qubits have T1 coherence times approaching 100 microseconds, about five times better than Google’s previous generation.[5][6] Its predecessor, the 54-qubit Sycamore, was the basis of Google’s October 2019 claim that a quantum processor had completed in 200 seconds a task it estimated would take a supercomputer 10,000 years.[7]

    Below-threshold error correction

    Willow’s headline result was error correction that improves with scale. Google ran a surface-code on grids of 3x3, 5x5 and 7x7 encoded qubits and found the logical error rate halved at each step.[2] The Nature paper reported 0.143% error per cycle on the 101-qubit distance-7 code, with a logical memory lasting about 2.4 times longer than the chip’s best physical qubit.[8][9] Decoding ran in real time, with an average decoder latency of 63 microseconds against a 1.1-microsecond cycle; Google called it one of the first compelling examples of real-time error correction on superconducting hardware.[10][11]

    Follow-up experiments (2026)

    In January 2026 Google reported dynamic surface code circuits on Willow, published in Nature Physics. Hexagonal circuits cut the number of couplers each qubit needs, walking circuits limit errors such as leakage, and iSWAP circuits allow a different two-qubit gate.[12] Simulating a three-coupler hexagonal layout by switching off unused couplers, Google found the logical error rate improved by a factor of 2.15 as the code grew from distance 3 to 5.[13][14] In July 2026 it reported in Nature a reinforcement-learning agent that learns from error-detection data and steers thousands of control parameters during a computation, which improved the code’s logical stability 3.5-fold.[15][16]

    Beyond-classical benchmarks

    At launch Google said Willow completed a random circuit sampling task in under five minutes that would take one of the fastest supercomputers 10 septillion years.[17] Google, which pioneered the benchmark, calls it the classically hardest test a quantum computer can run today and an entry point for checking whether a machine can beat classical computers at all.[18] In October 2025 Google reported in Nature its Quantum Echoes experiment on Willow, which measured out-of-time-order correlators in about two hours, a task it estimated would take 13,000 times longer on the Frontier supercomputer.[3] Google called this verifiable because the result is a physical quantity that another quantum computer or a natural quantum system can reproduce.[4] A companion study applied the method to molecular structure via nuclear magnetic resonance, but did not yet exceed classical simulation.[19]

    Significance and limits

    Willow put superconducting qubits at the centre of the error-correction race, but its results are memory and benchmark demonstrations, not useful computations. Google’s own stated next milestone is a useful, beyond-classical computation relevant to a real-world application.[20] Correlated errors occurring about once an hour remain a limit for larger codes.[21] Google researchers have also used Willow-class hardware assumptions to estimate that fewer than 500,000 physical qubits could break 256-bit elliptic-curve cryptography, far beyond Willow’s 105.[22][1] They published that estimate with a zero-knowledge proof so others could check it without the attack details being released.[23] See how quantum computers break encryption and how post-quantum cryptography works.

    Questions readers ask

    What did Willow prove?

    That error correction can improve as it scales. Each time Google grew its encoded grid from 3x3 to 5x5 to 7x7, the logical error rate halved.[2]

    Is Willow really 10 septillion years faster than a supercomputer?

    Only on random circuit sampling, a benchmark with no practical use. Google said Willow did that task in under five minutes versus an estimated 10 septillion years.[17]

    What is Quantum Echoes?

    An algorithm measuring out-of-time-order correlators that ran on Willow about 13,000 times faster than on the Frontier supercomputer, with outputs other quantum systems can verify.[3][4]

    Sources

    Each numbered claim is a statement we checked against the sources listed with it. Status shows how well established it is.

    1. [1]

      Google's Willow chip, announced on 9 December 2024, has 105 superconducting qubits. confirmedas of 2024-12-09

    2. [2]

      Google reported that as it scaled Willow's encoded qubit grids from 3x3 to 5x5 to 7x7, the logical error rate halved at each step, which it called the first below-threshold error correction. confirmedas of 2024-12-09

    3. [3]

      On 22 October 2025 Google reported in Nature that its Quantum Echoes (out-of-time-order correlator) algorithm ran on Willow in about two hours, a task it estimated would take 13,000 times longer on the Frontier supercomputer. confirmedas of 2025-10-22

    4. [4]

      Google called the Quantum Echoes result verifiable because its outputs are physical expectation values that another quantum computer or a natural quantum system can reproduce, unlike random bitstrings. confirmedas of 2025-10-22

    5. [5]

      Willow was fabricated in Google's dedicated quantum chip fabrication facility in Santa Barbara, California. confirmedas of 2024-12-09

    6. [6]

      Willow's qubits have T1 coherence times approaching 100 microseconds, about five times longer than Google's previous generation. confirmedas of 2024-12-09

    7. [7]

      On 23 October 2019 Google reported that its 54-qubit Sycamore processor performed a target computation in 200 seconds that it estimated would take the world's fastest supercomputer 10,000 years. confirmedas of 2019-10-23

    8. [8]

      Google's below-threshold experiment, published in Nature in 2025, ran a 101-qubit distance-7 surface code with 0.143% logical error per cycle, with errors suppressed by a factor of 2.14 each time the code distance grew by two. confirmedas of 2025-02-01

    9. [9]

      In the same experiment the distance-7 logical memory outlived the best physical qubit on the chip by a factor of about 2.4. confirmedas of 2025-02-01

    10. [10]

      The experiment decoded errors in real time, with an average decoder latency of 63 microseconds at distance 5 and an error-correction cycle time of 1.1 microseconds. confirmedas of 2025-02-01

    11. [11]

      Google described Willow's error-correction result as one of the first compelling examples of real-time error correction on a superconducting system, and as beyond breakeven because the encoded qubits outlived the physical ones. confirmedas of 2024-12-09

    12. [12]

      Google's dynamic surface code work, published in Nature Physics, demonstrated three circuit types - hexagonal circuits that reduce the number of couplers, walking circuits that limit errors such as leakage, and iSWAP circuits that allow non-standard two-qubit gates. confirmedas of 2026-01-13

    13. [13]

      On a hexagonal lattice each qubit would connect to three neighbours instead of four, which Google says would simplify the design and fabrication of large chips; it tested this on Willow by switching off unused couplers. confirmedas of 2026-01-13

    14. [14]

      In January 2026 Google reported dynamic surface code circuits that need fewer couplers and reduce correlated errors compared with static circuits, including a hexagonal variant that improved logical error rates by a factor of 2.15 as the code grew. confirmedas of 2026-01-13

    15. [15]

      Google's July 2026 Nature paper describes a reinforcement-learning agent that learns from error-detection data to steer thousands of control parameters during a computation, instead of halting the computation to recalibrate. confirmedas of 2026-07-22

    16. [16]

      In July 2026 Google reported that reinforcement-learning control, adjusting parameters during computation, improved the logical stability of its error-correcting code 3.5-fold on Willow. confirmedas of 2026-07-22

    17. [17]

      Google said Willow performed a random circuit sampling computation in under five minutes that would take one of the fastest supercomputers 10 septillion years. confirmedas of 2024-12-09

    18. [18]

      Google describes random circuit sampling, which it pioneered, as the classically hardest benchmark that can be run on a quantum computer today and an entry-point test of beating classical computers. confirmedas of 2024-12-09

    19. [19]

      A companion Google study applied the echo technique to nuclear magnetic resonance measurements of molecular structure, but those results were not yet beyond classical simulation. confirmedas of 2025-10-22

    20. [20]

      Google's stated next goal is a first useful, beyond-classical computation relevant to a real-world application, citing areas such as drug discovery, battery design and energy research. confirmedas of 2024-12-09

    21. [21]

      Repetition-code tests up to distance 29 found performance limited by rare correlated error events occurring about once an hour. confirmedas of 2025-02-01

    22. [22]

      In March 2026 Google researchers, in work framed around safeguarding cryptocurrency, estimated that 256-bit elliptic-curve cryptography could be broken with fewer than 1,200 logical qubits and 90 million Toffoli gates, or fewer than 500,000 physical superconducting qubits running for a few minutes, under standard hardware assumptions. confirmedas of 2026-03-31

    23. [23]

      Google said it used zero-knowledge proofs to let others verify its elliptic-curve resource estimate without publishing sensitive attack details. confirmedas of 2026-03-31

    Revision history (2)
    1. Page created.
    2. Added the Sycamore predecessor, real-time decoding context, details of the 2026 dynamic-circuit and reinforcement-learning experiments, and links to the encryption pages.

    Created Oct 10, 2026. Last reviewed by an editor on Oct 10, 2026. Next scheduled review: Jan 10, 2027.

    Cite this page

    "Google Willow." ContentLora, updated Oct 10, 2026. https://contentlora.com/wiki/google-willow

    Spotted an error? Suggest a correction or emailcorrections@contentlora.com.

    Keep exploring

    Connections
    Superconducting qub… Qubit technologySurface code Error-correcting code
    ☰ Contents