Do these 3 things before closing this tab:
1Scan for outdated or missing drivers - takes under a minute2Repair Windows errors before they cause bigger problems3Fix the driver behind crashes, sound loss and screen glitchesiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Google claims a quantum leap with its Willow chip, but the strongest result is narrower and more important than the headline suggests: Willow demonstrated below-threshold surface-code error correction, a prerequisite for scalable fault-tolerant quantum computing. Willow is not yet a general-purpose, fully fault-tolerant, or commercially available quantum computer.
Google introduced Willow on December 9, 2024, as a 105-qubit superconducting quantum processor. The processor produced two headline results: a random-circuit-sampling benchmark completed in under five minutes against Google’s estimate of 1025 years for a leading classical supercomputer, and a more consequential error-correction experiment in which larger encoded memories became more reliable.
The second result is the real technical story. Quantum processors are useful only if they can preserve information through long computations despite noisy hardware. Willow did not solve that problem, but Google reported an important step toward solving it.
Key takeaways
- Willow is a 105-physical-qubit superconducting processor; 105 physical qubits do not equal 105 reliable logical qubits.
- The Nature paper on Willow reported below-threshold surface-code behavior, including a distance-7 memory using 101 qubits and a logical error rate of 0.143% ± 0.003% per error-correction cycle.
- Google’s under-five-minute versus 1025-year comparison applies to random circuit sampling, a specialized benchmark rather than a normal business or scientific workload.
- Google’s October 22, 2025 Quantum Echoes announcement claimed a 13,000× advantage over a specified classical comparison, but the molecular experiments remained proof-of-principle research.
- Willow is not documented as a normal public cloud product with a posted per-job price; Google describes a selective, proposal-based Willow Early Access route.
What did Google’s Willow quantum chip actually prove?
Google’s Willow quantum chip demonstrated a meaningful advance in quantum-error correction, not a finished general-purpose quantum computer. The central result was that increasing the size of a surface-code memory reduced its logical error rate, indicating operation below the error-correction threshold for the tested experiment.
#1 Best Overall
That distinction matters because quantum processors are intrinsically noisy. A processor can contain many physical qubits and still be unable to run a long calculation if errors accumulate faster than the system can detect and correct them. Willow showed the scaling behavior needed by one leading error-correction approach: adding redundancy began improving the encoded memory instead of making it worse.
The result is a credible research milestone. The phrase “quantum leap” is reasonable shorthand for the error-correction engineering progress, but it becomes misleading if it implies that commercial quantum advantage, broad practical usefulness, or a fully fault-tolerant machine has already arrived.
What is Google Willow?
Willow is a superconducting quantum processor developed by Google Quantum AI. Google’s Willow specification sheet lists 105 qubits, average connectivity of 3.47—typically described as approximately four-way connectivity—and separate reported hardware configurations for error-correction experiments and random circuit sampling.
Quick wins for a faster PC:
Repair Windows errors before they cause bigger problemsFix Now →Scan for outdated or missing drivers - takes under a minuteDriver Scan →The reported specifications are laboratory hardware measurements, not a consumer performance rating. Depending on the configuration, the specification sheet reports single-qubit gate errors of approximately 0.035%–0.036%, two-qubit gate errors of 0.33% for the error-correction configuration and 0.14% for the random-circuit-sampling configuration, and mean T1 coherence times of 68 microseconds and 98 microseconds. The surface-code cycle time is reported as 1.1 microseconds.
| Willow characteristic | Reported value | Why it matters |
|---|---|---|
| Processor architecture | Superconducting quantum processor | Uses superconducting circuits operated in a cryogenic environment |
| Physical-qubit count | 105 | Raw hardware count; not a count of fully protected logical qubits |
| Average connectivity | 3.47 | Each qubit is connected to roughly four neighboring qubits on average |
| Single-qubit gate error | Approximately 0.035%–0.036% | Configuration-dependent measure of single-qubit operation fidelity |
| Two-qubit gate error | 0.33% for error correction; 0.14% for RCS | Different operating configurations have different trade-offs |
| Mean T1 coherence time | 68 or 98 microseconds, depending on configuration | Indicates how long an excitation tends to persist before relaxing |
| Surface-code cycle time | 1.1 microseconds | Time for one repeated error-correction cycle |
A raw qubit count is an incomplete way to compare quantum computers. Physical qubits are the noisy hardware elements. A logical qubit is an encoded unit of quantum information made from multiple physical qubits, together with repeated measurements and classical decoding. The engineering challenge is to obtain enough reliable logical qubits from a much larger pool of imperfect physical qubits.
Why does quantum-error correction matter more than the 105-qubit headline?
Quantum-error correction matters because useful algorithms may require operations lasting far longer than the coherence and gate-error limits of individual physical qubits. Error correction distributes one logical qubit across many physical qubits, measures information about errors, and uses a decoder to infer corrections without directly measuring away the computation’s quantum state.
Willow’s experiment used the surface code. In a surface-code memory, the number called the code distance describes the protection level and is associated with a larger arrangement of physical qubits. A higher distance generally requires more hardware but can make the encoded information harder to corrupt—provided the underlying physical error rate is low enough.
The threshold is the crucial dividing line. Above the threshold, adding more physical qubits can add more opportunities for failure and make the encoded result worse. Below the threshold, increasing code distance can reduce the logical error rate. Below-threshold performance is therefore a prerequisite for scaling, not the same thing as a completed fault-tolerant computer.
Google’s December 2024 explanation describes the core observation: moving from 3×3 to 5×5 to 7×7 layouts reduced the encoded error rate by roughly a factor of two at each step. The Google Quantum AI account of the Willow error-correction experiment presents that trend as the key milestone.
What did the Nature paper demonstrate?
The peer-reviewed Nature study, authored by Google Quantum AI researchers and collaborators, reported distance-5 and distance-7 surface-code memories with real-time decoding. The largest reported memory used 101 physical qubits in a distance-7 code.
The distance-7 memory had a logical error rate of 0.143% ± 0.003% per error-correction cycle. When the code distance increased by two, the paper reported an error-suppression factor of Λ = 2.14 ± 0.02. The paper also reported that the logical memory exceeded the lifetime of the best physical qubit by a factor of 2.4 ± 0.3, described as performance beyond breakeven.
The Tool Desk
Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Those figures answer several different questions:
- Does larger code size help? In the tested memories, yes. The logical error rate decreased as the code distance increased.
- Did the encoded memory beat the best underlying physical qubit? Google’s paper reported that it did, by the stated 2.4 ± 0.3 factor.
- Was the memory error-free? No. A 0.143% error rate per correction cycle is measurable and would compound over a sufficiently long computation.
- Could the classical decoder keep up? The paper reported decoder latency of 63 microseconds at distance 5, alongside real-time decoding operation.
- Was the machine fully fault tolerant? No. The paper identifies substantial further scaling as necessary for large-scale fault-tolerant algorithms.
The phrase “below threshold” should therefore be read precisely: Willow crossed an important operating regime for the tested surface-code memory. Willow did not produce a long-lived, general-purpose logical qubit capable of supporting arbitrary useful algorithms at commercial scale.
| Reported result | Value | Correct interpretation |
|---|---|---|
| Largest surface-code memory | Distance 7 using 101 qubits | A protected memory experiment, not 101 logical qubits |
| Logical error rate | 0.143% ± 0.003% per correction cycle | The encoded memory still experienced errors |
| Error suppression | Λ = 2.14 ± 0.02 for a distance increase of two | Larger tested codes suppressed errors by the reported factor |
| Beyond breakeven | 2.4 ± 0.3 relative to the best physical-qubit lifetime | The logical memory outperformed the comparison physical qubit |
| Decoder latency | 63 microseconds at distance 5 | Shows the importance of classical decoding speed in the system |
The Nature article includes an author correction published April 28, 2026. Readers relying on exact table values, equations, or interpretations should check the current version of the Nature article and its correction history before using the figures in technical work.
Does Willow really take five minutes instead of 1025 years?
Google reported that Willow completed a random circuit sampling benchmark in under five minutes, compared with Google’s estimate of 1025 years—10 septillion years in the short-scale U.S. naming system—for a leading classical supercomputer. The comparison is real as a benchmark claim, but it does not mean Willow is 1025 times faster at ordinary computing.
Random circuit sampling, or RCS, asks a quantum processor to generate samples from the output distribution of a deliberately difficult random quantum circuit. The task is useful for testing whether classical machines can efficiently simulate a particular quantum process. It is not the same as optimizing a supply chain, designing a drug, searching a database, rendering a video, or running a conventional software application.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
The Willow specification sheet attributes the 1025-year estimate to a classical simulation comparison. The estimate depends on the circuit, the simulation algorithm, the chosen classical hardware, and assumptions about the best available method. A different classical algorithm or benchmark definition could change the comparison.
| What the benchmark says | What it does not say |
|---|---|
| Willow generated samples for a specially designed random circuit in under five minutes. | Willow is 1025 times faster than a classical computer for every task. |
| The selected circuit is extremely difficult for classical simulation. | The benchmark has immediate commercial usefulness. |
| The result demonstrates a computational separation for that task. | Willow can replace classical computers. |
| Google estimated a classical runtime of 1025 years under its stated comparison. | 1025 years is a universal or final estimate for all classical simulation. |
The most accurate description is “a dramatic quantum advantage on a specialized random-circuit-sampling benchmark,” not “a general-purpose computer that is 1025 years faster.”
What changed with Google’s Quantum Echoes announcement?
On October 22, 2025, Google announced Quantum Echoes, a more application-oriented experiment on Willow. Google said the algorithm implemented an out-of-order time-correlator calculation, ran on Willow’s 105-qubit array, and was 13,000 times faster than the best classical algorithm in Google’s comparison.
Google also called the result “verifiable quantum advantage.” In this context, verifiable means that another sufficiently capable quantum computer could reproduce the result, rather than requiring trust in an uncheckable output. “Verifiable quantum advantage” is Google’s terminology and should not automatically be treated as a synonym for commercially established quantum advantage.
The announcement included proof-of-principle molecular experiments involving molecules with 15 and 28 atoms. Such work could eventually be relevant to molecular structure, materials science, or drug-discovery research. The experiments do not demonstrate an end-to-end drug-discovery service, a commercially useful molecular-design workflow, or a conclusive real-world advantage for industry.
The Google Quantum Echoes announcement supports a stronger claim than the RCS result in one respect: the workload is closer to a scientific measurement. The result remains an early research demonstration, and the claimed speedup is tied to Google’s specified classical algorithm and hardware comparison.
Is Willow a fully fault-tolerant quantum computer?
No. Willow’s below-threshold experiment is a step toward fault tolerance, not a demonstration of a fully fault-tolerant machine.
A practical fault-tolerant computer would need many logical qubits with sufficiently low error rates, reliable logical gates, sustained error correction over long computations, fast classical decoding, and an architecture that can scale without making cooling, control, wiring, and data movement unmanageable. Willow’s distance-7 memory showed improving protection, but it still had a nonzero logical error rate and used many physical qubits to protect one encoded memory.
Google’s next stated hardware milestone is a long-lived logical qubit. The unresolved scaling question is whether the error-suppression trend can continue through the much larger systems needed for useful fault-tolerant algorithms. More physical qubits alone will not answer that question; the quality and number of logical qubits matter more.
How should readers judge whether Willow is a breakthrough?
Willow is best evaluated using five separate tests rather than one headline number.
- Error suppression: Does logical reliability improve as more physical qubits are added? Google’s Willow experiment reported yes for the tested surface-code memories.
- Breakeven: Does an encoded logical qubit last longer than the physical qubits used to build it? The Nature paper reported performance beyond breakeven by 2.4 ± 0.3 against its physical-qubit comparison.
- Real-time operation: Can the decoder process error information quickly enough to support the processor? The Nature paper reported real-time decoding and a 63-microsecond decoder latency at distance 5.
- Algorithmic relevance: Does the processor perform a task connected to science or industry rather than only a synthetic benchmark? Quantum Echoes improves the relevance of the workload, but remains proof-of-principle.
- Reproducibility: Can another capable quantum system reproduce or verify the result? Google designed Quantum Echoes as a verifiable experiment, but that is not the same as broad independent replication.
On the first three tests, Willow’s reported results are strong. Quantum Echoes strengthens the fourth and fifth categories, while Google’s own later qualification remains important: in its November 2025 framework for useful quantum applications, Google said that no end-to-end quantum application with conclusive advantage on a real-world problem had yet been implemented in hardware.
What are the main trade-offs behind Willow?
Quantum hardware performance involves trade-offs that a single qubit count cannot capture.
PC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Crashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minute- Physical qubits versus logical reliability: More physical qubits can provide more error-correction redundancy, but they also increase control, cooling, wiring, decoding, and calibration demands.
- Qubit count versus qubit quality: A larger processor is not automatically better than a smaller processor with lower gate errors, stronger connectivity, and more stable operation.
- Coherence versus architecture: Willow’s two reported configurations have different coherence times and two-qubit error rates, showing that the chip does not have one universal performance number.
- Benchmark separation versus usefulness: A task can be far harder for a classical simulator without being commercially valuable.
- Quantum processing versus classical support: Useful quantum workloads still require classical compilation, orchestration, error decoding, data movement, and result validation.
- Research access versus commercial access: A selective research program can enable important experiments without functioning as an open, predictable cloud service.
Can the public buy or use Google Willow?
No ordinary consumer or developer can simply buy Willow, select it from a public cloud console, and run arbitrary workloads. The cited Google materials do not document a standard Willow purchase, a public per-shot price, or a normal self-service API.
Google’s public quantum-computing materials describe a Willow Early Access Program. The March 2026 Willow Early Access guidance describes a proposal process in which researchers submit experiments for evaluation, including consideration of technical feasibility and scientific impact. The program is therefore aimed at credible research groups rather than casual users seeking predictable pay-as-you-go access.
| Reader goal | Willow today | More practical route |
|---|---|---|
| Learn quantum programming | Not a normal public sign-up product | Use a public quantum SDK, simulator, or educational platform |
| Run a research experiment on Willow | Proposal-based Early Access route | Prepare a technically credible research proposal |
| Run jobs on public quantum hardware | No standard Willow price or open service documented | Consider IBM Quantum or Azure Quantum partner hardware |
| Use Google’s specific Willow architecture | Selective research access, not ordinary rental access | Monitor Google Quantum AI announcements and access guidance |
What can you use instead of Willow?
Readers who want hands-on quantum computing today generally need to use educational tools, simulators, or another provider’s public quantum hardware. These alternatives do not reproduce Willow’s architecture or Google’s error-correction experiment, but they can support learning and prototyping.
IBM Quantum
IBM Quantum provides access to IBM’s quantum processors and software ecosystem. IBM’s documentation describes an Open Plan with up to 10 minutes of quantum time per month at no charge, alongside paid enterprise and pay-as-you-go routes through IBM Cloud. Plan terms, hardware availability, queues, and access conditions can change, so readers should verify the current documentation before relying on the free allowance.
Free tools Windows power users keep installed
One-click scans. No signup required.
IBM is a sensible option for learners and developers who want direct access to a public QPU ecosystem and Qiskit tooling. IBM hardware is not Willow hardware, and IBM access does not reproduce Google’s surface-code experiment.
Azure Quantum
Azure Quantum provides a Microsoft Azure interface to partner providers including IonQ, Quantinuum, and Rigetti. Azure documents pay-as-you-go and subscription models, with billing varying by provider, device, execution method, and error-mitigation settings.
Azure’s current pricing documentation gives provider-specific examples rather than Willow pricing. The examples include minimum execution prices of $97.50 with error mitigation or $12.4166 without it for IonQ Aria 1, and $168.195 with error mitigation or $25.7899 without it for IonQ Forte 1. These prices are subject to change, apply to the specified provider configurations, and should not be generalized to all Azure Quantum jobs.
Azure Quantum is most useful for organizations already operating in Azure or teams that want several hardware vendors behind one cloud and billing environment. Azure Quantum is not a free, Willow-specific experience.
What must happen before Willow leads to useful quantum computing?
Google must turn a promising error-correction trend into a scalable system with many reliable logical qubits and algorithms whose total cost and performance beat the best classical alternatives for valuable problems.
- Demonstrate a long-lived logical qubit with substantially lower error rates over useful operating periods.
- Scale from a protected memory to many interacting logical qubits, not merely a larger count of physical qubits.
- Keep logical errors low during long computations, including logical gates and communication between encoded qubits.
- Develop fault-tolerant algorithms with practical resource requirements.
- Show end-to-end useful applications with transparent comparisons against strong classical methods.
- Obtain independent replication or verification from other capable quantum systems.
- Offer a predictable access and pricing model if the technology is to become a commercial service.
These requirements explain why below-threshold error correction matters so much. Without it, adding hardware does not reliably produce a better computer. With it, scaling becomes more plausible—but still technically and economically difficult.
Frequently Asked Questions
Is Google Willow 1025 times faster than a classical computer?
No. Google’s 1025-year comparison applies to a specialized random circuit sampling benchmark completed by Willow in under five minutes. The figure does not describe Willow’s speed on ordinary software, business workloads, or every quantum algorithm.
Does Willow have 105 logical qubits?
No. Willow is specified as a 105-physical-qubit processor. The reported error-correction experiment used many physical qubits to create a protected memory, including a distance-7 code using 101 qubits; that is not equivalent to 101 or 105 logical qubits.
Recommended Free Tools
Is Google Willow commercially available?
Willow is not documented in the cited official materials as a chip for sale or a normal public pay-per-use cloud service. Google describes a proposal-based Willow Early Access Program for researchers, with proposals evaluated for technical feasibility and scientific impact.
Did Quantum Echoes prove practical drug discovery on a quantum computer?
No. Google’s October 2025 Quantum Echoes announcement described a 13,000× advantage in a specified classical comparison and proof-of-principle molecular experiments involving 15- and 28-atom molecules. The work did not demonstrate an end-to-end commercial drug-discovery service or conclusive real-world advantage.
What is the most important Willow result?
The most important Willow result is the reported below-threshold surface-code error correction. Larger tested encoded memories became more reliable, which supports the possibility of scaling toward fault-tolerant quantum computing but does not establish that a fully fault-tolerant machine already exists.
The Bottom Line
Bottom line: Google Willow is a genuine and important quantum-error-correction milestone, not a finished commercial quantum computer. The under-five-minute result versus 1025 years is a specialized random-circuit-sampling benchmark, while Quantum Echoes is a more application-oriented but still early proof of principle. The decisive next test is whether Google can scale Willow’s below-threshold behavior into many long-lived logical qubits and demonstrate useful end-to-end applications.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
