Fact-checked by Grok 2 weeks ago

Sycamore processor

The Sycamore processor is a family of programmable superconducting quantum processors developed by Google Quantum AI, utilizing transmon qubits arranged in a two-dimensional lattice to perform complex quantum computations that surpass classical supercomputers in specific tasks.^[1] First introduced in 2019 with a 54-qubit (53 functional) configuration, it demonstrated quantum supremacy by sampling random quantum circuits in approximately 200 seconds—a feat estimated to require 10,000 years on the world's fastest classical supercomputer at the time.^[1]^[2] Subsequent iterations have scaled up the qubit count and improved fidelity, enabling experiments in entanglement phase transitions and random circuit sampling (RCS) benchmarks. In 2023, a 70-qubit version of Sycamore observed measurement-induced entanglement phase transitions in the largest system to date, revealing emergent quantum teleportation through random measurements and leveraging space-time duality to mitigate noise effects.^[3] By 2024, a 67-qubit Sycamore achieved an RCS fidelity of 1.5 × 10⁻³ over 32 cycles (880 two-qubit gates), operating in a low-noise regime that doubled circuit volume compared to 2019 while exceeding classical simulation capabilities, thus validating RCS as a scalable measure of quantum progress.^[4] In December 2024, Google announced Willow, a 105-qubit successor to Sycamore, achieving exponential error reduction and below-threshold performance; further advancements in 2025 included the Quantum Echoes algorithm demonstrating scalable quantum advantage.^[5]^[6] These advancements highlight Sycamore's role in pushing toward fault-tolerant quantum computing, with key technical features including high-fidelity gates (average two-qubit error of 0.62% in the original design), operation at millikelvin temperatures, and integration with adjustable couplers for tunable qubit interactions.^[1] Despite ongoing challenges like noise and error rates, Sycamore's evolution underscores Google's focus on hybrid quantum-classical algorithms for applications in optimization, simulation, and machine learning.^[3]^[4]

Development

Origins and initial project

Google established its Quantum Artificial Intelligence (Quantum AI) Lab in May 2013 as a collaborative effort with NASA and the Universities Space Research Association, initially focusing on leveraging quantum computing to advance machine learning algorithms.^[7] At launch, the lab utilized D-Wave's quantum annealing systems to explore potential applications in optimization problems relevant to artificial intelligence.^[8] By 2014, Google shifted its strategy toward developing in-house universal quantum computing hardware based on superconducting qubits, moving beyond the limitations of D-Wave's specialized annealing approach.^[9] This transition was spearheaded by the recruitment of physicist John Martinis and his team from the University of California, Santa Barbara, who brought expertise in fabricating high-fidelity superconducting quantum circuits.^[10] Martinis established a dedicated quantum hardware lab near Santa Barbara to pursue scalable quantum processors.^[11] The initial objectives of the project centered on achieving quantum supremacy, defined as demonstrating a quantum computation that outperforms the capabilities of classical supercomputers for a specific task, thereby validating the potential of quantum hardware.^[10] Development of the core technology began in earnest around 2015–2016, with a strong emphasis on transmon qubits—superconducting circuits designed for improved coherence times and scalability to larger arrays.^[12] This focus enabled iterative progress toward fault-tolerant systems capable of error-corrected operations.^[9]

Fabrication and key milestones

The Sycamore processor employs superconducting transmon qubits, which are fabricated on high-resistivity silicon wafers using aluminum for the circuits and Josephson junctions. The fabrication process incorporates 14 lithography layers to create the qubit and coupler structures, followed by assembly through indium bump bonding of two silicon dies to enable 3D integration and low-loss wiring. This design allows for tunable coupling between qubits via frequency control, essential for implementing the required quantum gates.^[1] To maintain quantum coherence, the processor operates at temperatures below 20 millikelvin, achieved using dilution refrigerators that feature a mixing chamber stage for precise cooling and cryogenic amplifiers to handle control signals with minimal added noise. These extreme conditions suppress thermal excitations that could disrupt the superconducting state of the qubits, which resonate at frequencies between 5 and 7 GHz.^[1] Development milestones for Sycamore included prototype testing of key components in 2018, building on prior Google quantum hardware like the Bristlecone processor, followed by the complete 53-qubit assembly and calibration in early 2019, where one qubit from the intended 54-qubit array proved non-functional due to fabrication defects.^[13]^[1] A major engineering challenge was managing qubit decoherence, with median energy relaxation times (T1) around 20 microseconds at operating frequencies—well under 100 microseconds—necessitating careful frequency tuning and error mitigation strategies to enable reliable circuit execution over multiple gate layers.^[1]^[14]

Technical design

Qubit architecture

The Sycamore processor employs transmon qubits, which are superconducting quantum bits realized as nonlinear resonators operating at frequencies between 5 and 7 GHz, with quantum information encoded in their two lowest eigenstates.^[1] These transmons incorporate Josephson junctions to provide an anharmonicity that enables selective addressing of qubit states, and they were selected for their reduced sensitivity to charge noise compared to earlier charge-based superconducting qubits, allowing for longer coherence times and higher gate fidelities.^[1] Key parameters of the Sycamore transmon qubits include coherence times on the order of 20–50 microseconds, with typical energy relaxation time T_1 \approx 20 \, \mu\text{s} and dephasing time T_{2,\text{CPMG}} \approx 30 \, \mu\text{s}, enabling the execution of deep quantum circuits before decoherence significantly impacts performance.^[15] Single-qubit gate fidelities exceed 99.9%, achieved through microwave pulse drives resonant with the qubit frequency, corresponding to an average Pauli error rate of about 0.1%, while two-qubit gate fidelities are approximately 99.4–99.6% (errors of 0.36–0.62%).^[1]^[16] Error rates in the Sycamore qubits are dominated by two-qubit operations, with an average gate error of 0.2–0.6%; specifically, isolated two-qubit gates exhibit 0.36% error, rising to 0.62% under simultaneous multi-qubit operation due to crosstalk and control imperfections, which contribute to overall circuit-level noise.^[1] These errors arise primarily from residual coupling during gates and relaxation during idling, but they remain low enough for short-depth computations. For scaling to larger systems, the Sycamore design incorporates 86 tunable couplers alongside its 54 transmon qubits (53 of which are active), allowing the inter-qubit coupling strength to be rapidly adjusted from zero to 40 MHz.^[1] This tunability suppresses unwanted interactions during single-qubit operations and enables high-fidelity entangling gates, facilitating connectivity in a 2D lattice suitable for surface code error correction in future iterations.^[1]

Processor layout and operations

The Sycamore processor consists of 54 superconducting transmon qubits arranged in a two-dimensional rectangular grid, with 53 qubits actively used in operations due to one defective qubit.^[1] This layout enables a scalable architecture where each qubit connects to up to four nearest neighbors, facilitating efficient implementation of quantum circuits on a planar surface. The connectivity is achieved through 86 tunable superconducting couplers positioned between adjacent qubits, which allow precise control over inter-qubit interactions by adjusting the coupling strength from near zero (off state) to approximately 40 MHz (on state).^[1] These couplers, implemented as additional transmon-like devices, mediate capacitive coupling and suppress unwanted crosstalk, enabling high-fidelity nearest-neighbor gates without requiring long-range connections. The processor supports a native gate set that enables universal quantum computation, comprising single-qubit rotations and two-qubit entangling operations. Single-qubit gates are realized using 25-nanosecond microwave pulses to drive XY rotations on the Bloch sphere, achieving fidelities exceeding 99.9%.^[1] For two-qubit interactions, the primary gates are the iSWAP (a partial swap operation with a duration of about 12 nanoseconds at 20 MHz coupling) and the fSim gate, which generalizes iSWAP by incorporating an additional controlled-phase (CZ) interaction, typically parameterized as fSim(π/2, π/6) for the supremacy experiments. The fSim gate, with its Schmidt rank of 4, provides a compact representation for entangling operations and can be decomposed into iSWAP plus a π/6 CZ phase, supporting the generation of arbitrary two-qubit unitaries when combined with single-qubit gates.^[1] Overall two-qubit gate fidelities reach approximately 99.4% on average across the device. Control of the processor relies on a sophisticated cryogenic electronics system operating at millikelvin temperatures. Microwave pulses for single-qubit manipulations are generated by 54 independent arbitrary waveform generators, each sampling at 1 gigasample per second and phase-synchronized to ensure coherent drive across the array. Flux tuning, achieved via fast digital-to-analog converters connected to on-chip bias lines, adjusts the frequencies of both qubits (over a range of about 100 MHz) and couplers to enable or disable interactions dynamically during gate execution.^[1] This flux-based control allows rapid switching between idling and operating states, minimizing decoherence and supporting the execution of deep circuits. In terms of operational depth, the Sycamore processor demonstrates the capability to execute random quantum circuits with up to 20 layers of two-qubit gates on 53 qubits, corresponding to over 1,100 single-qubit and 430 two-qubit operations in the supremacy benchmark while maintaining a circuit fidelity above 0.1%.^[1] This depth highlights the device's suitability for complex computations, limited primarily by cumulative error rates rather than architectural constraints.

Quantum supremacy claim

The 2019 experiment

The 2019 experiment on the Sycamore processor aimed to demonstrate quantum supremacy through random quantum circuit sampling (RQCS), a computational task designed to produce samples from the output probability distribution of complex quantum circuits that are expected to be intractable for classical computers. The specific RQCS task involved generating bitstrings from random 53-qubit circuits with a depth of 20 cycles, where each cycle alternated layers of single-qubit gates (chosen randomly from the set {√X, √Y, √W}, where W = (X + Y)/√2) and two-qubit entangling gates (specifically, the fSim gate) applied to nearest-neighbor pairs in the 2D lattice, tiled across four subsets (ABCD) for simultaneous execution, followed by a final layer of single-qubit gates before measurement in the computational basis. This setup created a highly entangled quantum state whose output distribution exhibits speckled interference patterns due to quantum superposition, making exact classical simulation exponentially resource-intensive as the number of qubits and depth increase.^[1] The experimental setup utilized the 53 functional qubits of the Sycamore processor, configured in a 2D grid layout with tunable couplers for precise control. To acquire the necessary data for the supremacy demonstration and fidelity verification, the team executed multiple instances of these random circuits, collecting a total of 30 million output samples for the 53-qubit, depth-20 case across ten distinct circuit instances; this process spanned several days to account for device recalibration, error mitigation, and statistical robustness. Each circuit instance was sampled repeatedly on the processor, with outputs recorded as bitstrings to approximate the ideal probability distribution.^[1] As a benchmark for classical infeasibility, the experiment compared Sycamore's performance to the Summit supercomputer, then the world's fastest, estimating that the latter would require approximately 10,000 years to generate an equivalent set of samples with comparable fidelity (this figure was based on extrapolated simulations for smaller circuits and tensor network methods, though it was later revised following external analyses).^[1] Verification of the experiment's outputs relied on cross-entropy benchmarking (XEB), a fidelity metric that quantifies agreement between the experimentally observed bitstring frequencies and the ideal probabilities computed via classical simulation for the same circuits. XEB is defined as F_{\text{XEB}} = 2^n \langle P(x_i) \rangle_i - 1, where n is the number of qubits, P(x_i) are the ideal probabilities for the measured bitstrings x_i, and the average is over the experimental samples; values significantly exceeding the classical limit of 1 confirm a non-classical output distribution, as random classical sampling would yield F_{\text{XEB}} \approx 1. For the 53-qubit circuits, this method achieved F_{\text{XEB}} = (2.24 \pm 0.21) \times 10^{-3}, providing strong evidence that the samples originated from genuine quantum interference rather than classical noise or approximation.^[1]

Performance metrics

In the 2019 quantum supremacy experiment, the Sycamore processor completed the random quantum circuit sampling (RQCS) task in approximately 200 seconds, generating 1 million output bitstrings from a 53-qubit circuit with 20 cycles (with a total of 30 million samples collected across instances for verification).^[1] This runtime reflects the net time for sampling after accounting for processor overhead, highlighting the processor's efficiency in producing quantum states beyond classical reach for the full task.^[1] The experiment achieved a linear cross-entropy benchmarking (XEB) fidelity of (2.24 \pm 0.21) \times 10^{-3} (or about 0.2%) for the full 53-qubit, 20-cycle circuits, serving as a measure of how closely the quantum outputs matched ideal distributions.^[1] For partial samplings involving fewer cycles or qubits, the fidelity was higher, reaching up to several percent, which allowed for validation against classical simulations in those regimes.^[1] The overall predicted circuit fidelity for the largest instances, incorporating 1,113 single-qubit gates and 430 two-qubit gates, was approximately 0.2%, underscoring the challenges of noise in scalable quantum systems.^[1] Google's initial assessment estimated that simulating the full sampling task with comparable fidelity on classical hardware, such as a supercomputer with a million cores akin to Summit's capabilities, would require about 10,000 years.^[1] IBM contested this, proposing optimizations like output distillation and improved memory usage that could complete a high-fidelity simulation on the Summit supercomputer in 2.5 days.^[17] These benchmarks illustrate the tight competition between quantum and classical resources at the time. A key scalability metric from the experiment is the exponential resource growth in classical simulations beyond 50 qubits, where the computational cost scales factorially with circuit depth and qubit count, rendering full emulation infeasible on current supercomputers without approximations.^[1] This slowdown emphasizes why the 53-qubit threshold marked a significant, albeit specific, demonstration of quantum advantage in sampling tasks.^[1]

Controversies

IBM's rebuttal

In response to Google's announcement of quantum supremacy using the Sycamore processor, IBM published a blog post on October 21, 2019, challenging the claim by demonstrating that the experiment could be feasibly simulated on classical hardware. IBM asserted that their supercomputer Summit could perform an ideal simulation of the same task in 2.5 days using disk storage, achieving far greater fidelity than Google's reported 0.2% for the quantum circuit. This contradicted Google's estimate of 10,000 years for a classical supercomputer to complete the simulation, highlighting IBM's view that the task did not truly demonstrate supremacy over classical computing.^[17] IBM critiqued Google's methodological assumptions, particularly the underestimation of classical simulation efficiency due to the circuit's relatively low depth of 20 cycles, comprising 430 two-qubit gates and 1,113 single-qubit gates, which limited the complexity exploitable by quantum hardware. The company argued that Google's Schrödinger-Feynman simulation approach overlooked practical optimizations, such as using disk storage for intermediate data and techniques like circuit partitioning and deferred tensor contractions, which could dramatically reduce computational demands. Additionally, IBM emphasized that the experiment's focus on output probability sampling—rather than full state vector computation—made it particularly amenable to classical verification, as only a subset of the output distribution needed to be sampled for validation.^[17] As a direct counter-demonstration, IBM proposed deploying their forthcoming 53-qubit Stretch processor to execute a more industrially relevant task that would showcase quantum advantage without the vulnerabilities of Google's random circuit sampling. This processor, planned for release soon after the blog post, was positioned as a platform to advance verifiable quantum utility in areas like chemistry simulations, underscoring IBM's emphasis on practical applications over abstract supremacy benchmarks.^[17]

Broader scientific responses

The Sycamore processor's quantum supremacy demonstration elicited endorsements from prominent physicists, who viewed it as a significant milestone in showcasing non-trivial quantum computational advantages over classical systems. Scott Aaronson, a leading quantum computing theorist, described the achievement as a "huge and very welcome milestone" for the field, emphasizing its role in validating theoretical predictions about quantum speedup for specific sampling tasks.^[18] Skepticism arose particularly regarding the implications of the "supremacy" terminology, which some critics argued overstated practical utility and downplayed persistent challenges like noise in near-term quantum devices. Mathematician Gil Kalai raised concerns about the experiment's noise levels, suggesting through statistical analysis that the output distributions might not reliably demonstrate genuine quantum behavior due to potential classical correlations or data artifacts.^[19] In response to such debates, the scientific community increasingly favored "quantum advantage" as a more precise and less hyperbolic term, focusing on task-specific speedups rather than broad superiority. A 2020 Nature commentary highlighted this shift, noting how subsequent challenges to Sycamore's claims, such as those from photonic quantum experiments, underscored the need for clearer definitions to advance verifiable progress.^[20] Ongoing controversies have persisted with advances in classical simulation techniques. In 2024, researchers from the University of Science and Technology of China demonstrated an energy-efficient classical algorithm using 1,432 GPUs to simulate quantum random circuit sampling tasks similar to those performed by Sycamore, achieving results 7 times faster and with lower energy consumption than the 2019 quantum processor. This work further questioned the scalability of Sycamore's supremacy claims by showing classical methods can now outperform early quantum benchmarks in specific metrics.^[21] The claim spurred broader impacts, including heightened focus on verifying quantum simulations and accelerating research efforts in error mitigation and benchmarking protocols. This momentum contributed to expanded investigations into quantum verification techniques, fostering collaborations aimed at distinguishing true quantum effects from classical simulability.^[22]

Subsequent uses and legacy

Advanced simulations

Following the 2019 quantum supremacy demonstration, the Sycamore processor and its upgraded variants were employed in a series of advanced scientific simulations, leveraging increased qubit counts and improved coherence times to tackle problems in quantum chemistry and condensed matter physics. In 2020, researchers at Google Quantum AI used the 53-qubit Sycamore to perform the largest chemical simulation on a quantum computer to date, computing Hartree-Fock approximations for molecular energy levels in diazene (N₂H₂) and stretched hydrogen chains, achieving results that aligned with classical benchmarks while demonstrating quantum advantage in scaling to larger systems.^[23] In 2022, simulations of more challenging correlated molecules, such as the electronic structure of iron-sulfur clusters relevant to nitrogenase enzymes, were performed using variational quantum eigensolver algorithms on a Sycamore-derived processor with up to 9 qubits to capture static and dynamical properties beyond classical tractability for these systems.^[24] A landmark achievement in 2021 was the first experimental realization of a discrete time crystal on the Sycamore processor, where a 20-qubit array was driven periodically to exhibit perpetual oscillations without energy input, confirming a novel non-equilibrium phase of matter predicted by Floquet theory and inaccessible to classical equilibrium simulations.^[25] This work established time crystals as a verifiable quantum phase, with subharmonic response persisting over hundreds of cycles despite noise, and opened pathways to studying exotic matter phases like Floquet insulators. In 2023, to enhance simulation fidelity for longer computations, Google demonstrated surface code error correction on its 3rd-generation 70-qubit Sycamore processor, encoding a single logical qubit across 49 physical qubits and achieving logical error rates below the physical error threshold while suppressing errors by over 2.14 times compared to uncorrected qubits.^[15] This milestone enabled more robust simulations by mitigating decoherence, directly supporting subsequent experiments in non-equilibrium dynamics. Building on these advances, in 2025, a 72-qubit Sycamore-class processor was used to probe Floquet topological order, realizing a driven quantum many-body system that exhibited chiral edge modes and bulk-boundary correspondence in a non-equilibrium phase of matter, which classical computers cannot efficiently simulate due to the exponential growth of Hilbert space under periodic driving (as of September 2025).^[26] These simulations provided the first experimental evidence of such topological phases, revealing anyon-like excitations and braiding statistics essential for fault-tolerant quantum computing, while highlighting Sycamore's role in accessing exotic matter states previously confined to theory.

Influence on quantum computing

The Sycamore processor marked a pivotal advancement in superconducting quantum hardware, serving as a foundational platform for subsequent iterations aimed at error correction. In 2023, Google upgraded Sycamore to a 70-qubit configuration, enabling demonstrations of quantum error correction with logical qubits encoded using surface codes on up to 49 physical qubits.^[15] This version contributed significantly to Google's quantum scaling roadmap, which outlines milestones toward high-fidelity, fault-tolerant systems by improving qubit connectivity and gate fidelities through iterative refinements in transmon qubit design.^[27]^[28]^[29] Sycamore's achievements accelerated progress across the quantum computing field, prompting competitors to enhance their superconducting technologies. For instance, IBM's 2021 release of the 127-qubit Eagle processor was part of a broader push to scale beyond Sycamore's 53-qubit supremacy demonstration, focusing on modular architectures to achieve similar milestones in coherence and gate performance. Similarly, Rigetti Computing advanced its multi-chip superconducting systems, achieving faster gate speeds over 1,000 times those of alternative modalities, in response to the performance benchmarks set by Sycamore's tunable couplers and fSim gates. These developments underscored Sycamore's role in fostering a competitive ecosystem for scalable quantum processors.^[17]^[30] The processor's influence extended to its direct successors, exemplifying evolutionary hardware design. In late 2024, Google transitioned to the Willow chip with 105 qubits, which built upon Sycamore's fSim entangling gates to achieve two-qubit error rates below the surface code threshold, enabling exponential error reduction as code distances increased. Willow's below-threshold performance on distance-7 and distance-5 codes represented a key step forward from Sycamore's error mitigation techniques. Looking ahead, Sycamore stands as a proof-of-concept for the million-qubit-scale systems required for practical fault-tolerant quantum computing, validating the viability of superconducting approaches in overcoming noise barriers en route to useful applications (as of November 2025).^[5]^[31]^[32]^[33]

References

[1]
Quantum supremacy using a programmable superconducting ...
Oct 23, 2019 · Our Sycamore processor takes about 200 seconds to sample one instance of a quantum circuit a million times—our benchmarks currently indicate ...A Suitable Computational... · Building A High-Fidelity... · Fidelity Estimation In The...
[2]
Quantum Supremacy Using a Programmable Superconducting ...
Oct 23, 2019 · We developed a new 54-qubit processor, named “Sycamore”, that is comprised of fast, high-fidelity quantum logic gates, in order to perform the benchmark ...
[3]
Measurement-induced entanglement phase transitions in a quantum ...
Oct 18, 2023 · ... 70 qubits on our Sycamore quantum processor. This is, by far, the largest system in which such a phase transition has been observed ...
[4]
Validating random circuit sampling as a benchmark for measuring ...
Oct 10, 2024 · This experiment presents new RCS results with an estimated fidelity of 1.5 × 10-3 at 67 qubits and 32 cycles (circuit depth) or 880 two-qubit ...
[5]
Launching the Quantum Artificial Intelligence Lab - Google Research
Launching the Quantum Artificial Intelligence Lab. May 16, 2013. Posted by Hartmut Neven, Director of Engineering. Quick links. Share. Copy link.Missing: division establishment
[6]
Google and NASA Launch Quantum Computing AI Lab
The new Quantum Artificial Intelligence Lab will employ what may be the most advanced commercially available quantum computer, the D-Wave Two.
[7]
Google branches out from D-Wave in quantum computing initiative
Sep 5, 2014 · Google branches out from D-Wave in quantum computing initiative. Working with academic to make more traditional superconducting qubits. John ...Missing: shift | Show results with:shift
[8]
The Man Who Will Build Google's Elusive Quantum Computer | WIRED
Sep 5, 2014 · John Martinis is one of the world's foremost experts on quantum computing, a growing field of science that aims to process information at super high speeds.
[9]
Google Launches Effort to Build Its Own Quantum Computer
Sep 3, 2014 · Now John Martinis, a professor at University of California, Santa Barbara, has joined Google to establish a new quantum hardware lab near ...
[10]
Google Hires Quantum Computing Expert John Martinis to Build ...
Google's new project has recruited an outside expert with a very different mindset for transforming quantum computing into a practical technology.
[11]
Assessing the Quantum-Computing Landscape
Oct 1, 2022 · In 2018, Google unveiled Bristle-cone, its 72-qubit processor.<|control11|><|separator|>
[12]
https://spectrum.ieee.org/google-hires-quantum-computing-expert-john-martinis-to-build-new-hardware
[13]
Suppressing quantum errors by scaling a surface code logical qubit
Feb 22, 2023 · Each qubit is coupled to four nearest neighbours except on the boundaries, with mean qubit coherence times T1 = 20 μs and T2,CPMG = 30 μs, in ...Surface Codes With... · Author Information · Google Quantum Ai
[14]
Optimal readout of superconducting qubits exploiting high-level states
We present a theoretical tool to extract different types of errors in high-level states encoding dispersive measurement.
[15]
On “quantum supremacy” | IBM Quantum Computing Blog
Oct 22, 2019 · The concept of “quantum supremacy” showcases the resources unique to quantum computers, such as direct access to entanglement and superposition.
[16]
What Google's Quantum Supremacy Claim Means ... - IEEE Spectrum
Sep 27, 2019 · “For those of us who work in quantum computing, the achievement of quantum supremacy is a huge and very welcome milestone,” says Scott Aaronson, ...
[17]
[PDF] Questions and Concerns About Google's Quantum Supremacy Claim
May 1, 2023 · (Recent improved classical algorithms have largely refuted Google's quantum supremacy claim and therefore the Google results no longer refute ...
[18]
Physicists in China challenge Google's 'quantum advantage' - Nature
Dec 3, 2020 · Photon-based quantum computer does a calculation that ordinary computers might never be able to do. By. Philip Ball. Philip Ball. View author ...
[19]
Google and NASA Achieve Quantum Supremacy
Oct 23, 2019 · The paper describes the experiments run by Google's Sycamore quantum processor to demonstrate quantum supremacy. Computations on a quantum ...
[20]
Simulating Models of Challenging Correlated Molecules and ...
Nov 14, 2022 · We simulate static and dynamical electronic structure on a superconducting quantum processor derived from Google's Sycamore architecture for two representative ...
[21]
Time-crystalline eigenstate order on a quantum processor - Nature
Nov 30, 2021 · In an equilibrium setting, quantum phases of matter are classified by long-range order or broken symmetries in low-temperature states (Fig. 1a).
[22]
Probing non-equilibrium topological order on a quantum processor
Sep 10, 2025 · Here we realize a Floquet topologically ordered state theoretically proposed in ref., on an array of superconducting qubits. We image the ...
[23]
Google's improved quantum processor good enough for error ...
Feb 22, 2023 · Today, Google announced a demonstration of quantum error correction on its next generation of quantum processors, Sycamore.
[24]
Google Quantum Computer: Shaping Tomorrow's Technology
Feb 24, 2025 · As of July 2023, Google has upgraded Sycamore to 70 qubits. In a recent experiment, this 70-qubit processor completed a computational task in ...Quantum Computing In A... · Google's Quantum Leap... · The Google Quantum Computer...<|control11|><|separator|>
[25]
[PDF] Our focused and responsible approach to quantum computing
Google Quantum AI aims to build this computer. Our roadmap is shown ... Quantum supremacy using a programmable superconducting processor. Nature 574 ...
[26]
Rigetti Demonstrates Industry's Largest Multi-Chip Quantum Computer
Jul 16, 2025 · “We benefit from the many advantages of superconducting qubits, including gate speeds more than 1,000x faster than other modalities like ion ...Missing: Sycamore | Show results with:Sycamore
[27]
Meet Willow, our state-of-the-art quantum chip - The Keyword
Dec 9, 2024 · Our new quantum chip demonstrates error correction and performance that paves the way to a useful, large-scale quantum computer.
[28]
Quantum error correction below the surface code threshold - Nature
Dec 9, 2024 · We begin with results from our 105-qubit Willow processor depicted in Fig. 1a. It features a square grid of superconducting transmon qubits ...
[29]
Google Debuts New Quantum Chip, Error Correction Breakthrough ...
Dec 9, 2024 · “On the performance front, importantly, we've been able to increase our qubit coherence times, the t1 value, by a factor of five, from 20 up to ...
[30]
The Google Quantum Computing Chip Willow: An Inside Look
Sep 29, 2025 · Willow is Google Quantum AI's newest superconducting quantum computing chip, developed as the successor to the company's groundbreaking Sycamore ...