⚛️ quantum physics

Modeling Logical Gates and Read-Out of Superconducting Gottesman-Kitaev-Preskill Qubits

This paper bridges the gap between theoretical GKP error correction models and practical superconducting device limitations by demonstrating that loss and dephasing can be approximated as random displacements, proposing decoder modifications to drastically reduce gate infidelity, and outlining a quantum trajectory-based scheme to mitigate measurement inefficiencies during read-out.

Original authors: Mackenzie H. Shaw, Andrew C. Doherty, Arne L. Grimsmo

Published 2026-10-01

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Mackenzie H. Shaw, Andrew C. Doherty, Arne L. Grimsmo

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

To build a computer that can solve problems beyond the reach of today's machines, scientists must first solve a fundamental problem: quantum information is incredibly fragile. The bits of data in a quantum computer, called qubits, tend to lose their state the moment they interact with their environment. To protect this delicate information, researchers use a strategy called quantum error correction. Instead of relying on a single, perfect physical component, they encode a single piece of logical information across a larger system, allowing them to detect and fix mistakes without destroying the data. One of the most promising ways to do this involves using bosonic codes, which store information in the continuous vibrations of a quantum oscillator, much like a pendulum swinging back and forth. Among these, the Gottesman-Kitaev-Preskill, or GKP, code stands out because it offers a path to universal quantum computing using relatively simple resources. However, a significant gap has existed between the theoretical models used to design these codes and the messy reality of building them in a laboratory. Theories often assume a specific, uniform type of noise, while real-world devices suffer from energy loss, random jitters, and imperfect measurements.

A team of researchers has now bridged this gap by developing new methods to model how GKP qubits behave under realistic conditions. They focused on three critical areas: how the qubits degrade over time, how logical operations spread errors, and how to read the final result when the measurement tools are imperfect. Their work reveals that while the theoretical promise of GKP codes remains strong, the practical path to a working quantum computer requires careful adjustments to how errors are corrected and how measurements are performed. By refining the way scientists decode the state of these qubits and by proposing new ways to boost the efficiency of their measurements, the researchers show that it is possible to significantly reduce the error rates that currently threaten to derail these experiments.

The researchers began by addressing the disconnect between the idealized noise models used in simulations and the actual physical processes that degrade GKP qubits in superconducting circuits. In theory, these codes are designed to correct uniform random shifts in position and momentum. In practice, the qubits suffer from energy loss and dephasing, which are more complex and difficult to simulate. The team developed a mathematical technique to approximate the effect of these real-world noises as a random shift, similar to the idealized model. This approximation allowed them to run efficient simulations that matched the results of more complex, exact calculations. They found that this method works remarkably well, providing a reliable way to predict how the quality of the qubit state, defined by its energy and squeezing, influences its ability to resist errors. This confirmation is vital because it allows researchers to use simpler, faster models to design better error-correction strategies without sacrificing accuracy.

Next, the team investigated what happens when logical gates, the operations that manipulate the qubits to perform calculations, are applied. They discovered that standard gates tend to spread errors across the system, making them harder to correct and reducing the overall fidelity of the operation. To counter this, they proposed a modification to the error-correction process itself. Instead of using a fixed, standard method to decode the qubit's state after every gate, they showed that the decoding strategy could be dynamically adjusted based on the specific gate that was just performed. By changing the shape of the region used to interpret the measurement results, they could effectively cancel out the error-spreading caused by the gate. In their simulations, this simple adjustment reduced the error rate of these operations by up to two orders of magnitude, a massive improvement that brings the performance of these gates much closer to the levels required for practical computing.

The final piece of the puzzle involved the read-out process, where the quantum state is measured to reveal the result of a calculation. In superconducting circuits, this is typically done using homodyne detection, a technique that measures the oscillation of the electromagnetic field. However, current experimental setups are far from perfect, with measurement efficiencies often ranging between 60 and 75 percent. This inefficiency introduces significant errors, making it difficult to distinguish the true state of the qubit from noise. The researchers analyzed the impact of this inefficiency and found that even with the best current hardware, the error rate would remain too high for reliable computing. To solve this, they proposed two schemes to improve the effective efficiency of the measurement. The first involves amplifying the signal before it is measured, while the second uses a clever coupling technique to transfer the information from the fragile GKP qubit to a separate, auxiliary mode that is easier to measure.

Using the theory of quantum trajectories, which tracks how a system evolves under continuous observation, the team demonstrated that these schemes could dramatically improve the measurement quality. They showed that by coupling the main qubit to a low-quality auxiliary mode and measuring that mode, they could achieve an effective measurement efficiency of over 90 percent. This level of efficiency is sufficient to bring the measurement error down to the 0.1 percent range, a threshold necessary for fault-tolerant quantum computing. Their analysis also highlighted that the speed of this measurement is crucial; if the process takes too long, the qubit will lose its state to natural decay. They calculated that with realistic coupling strengths, the measurement could be completed in less than a microsecond, a timeframe comparable to the read-out speeds of other leading quantum technologies.

The study concludes that while the path to a large-scale quantum computer using GKP codes is fraught with challenges, the specific hurdles of noise modeling, gate-induced errors, and measurement inefficiency are surmountable. By refining the theoretical models to match reality and introducing adaptive decoding and enhanced measurement techniques, the researchers have provided a concrete roadmap for improving the performance of these systems. Their work suggests that with the right engineering adjustments, the GKP code can fulfill its promise as a robust foundation for the next generation of quantum computers. The findings do not claim that the problem is solved, but rather that the tools to solve it are now within reach, offering a clearer view of how to move from theoretical possibility to experimental reality.

Technical Summary: Modeling Logical Gates and Read-Out of Superconducting Gottesman-Kitaev-Preskill Qubits

Problem Statement
The Gottesman-Kitaev-Preskill (GKP) code offers a promising pathway to fault-tolerant quantum computing, as universal computation can be achieved using Gaussian resources combined with GKP Pauli-eigenstate preparation. However, a significant disconnect exists between the theoretical noise models used to analyze GKP codes and the practical conditions of superconducting devices. Theoretical studies often rely on idealized uniform random displacement errors, whereas experimental implementations face realistic noise channels (loss and dephasing), imperfect logical gates, and inefficient homodyne measurements (typically 60–75% efficiency). Furthermore, existing multi-mode simulations often use incompatible noise models compared to single-mode studies, and the specific effects of Clifford gates on error spreading and measurement inefficiencies on logical read-out have not been fully bridged in a unified framework.

Methodology
The authors address these gaps by integrating three distinct theoretical techniques:

Stabilizer Subsystem Decomposition (SSSD): Utilizing the SSSD framework, the authors analytically estimate gate infidelities and noise effects. This method allows the partial trace over the stabilizer subsystem to correspond exactly to an ideal decoding map, facilitating the calculation of gate fidelities and error spreading without the high computational cost of traditional Fock space simulations as the GKP codestate quality improves ( $\Delta \to 0$ ).
Twirling Approximation Generalization: The authors derive a random displacement channel approximation for loss and dephasing acting on approximate GKP codestates. This generalizes the "twirling approximation" previously limited to envelope operators, showing that the effect of loss and dephasing can be modeled as a Gaussian random displacement channel with a specific variance derived from the envelope and noise parameters.
Quantum Trajectories: To analyze the impact of inefficient homodyne measurements, the authors employ quantum trajectory theory. They model the evolution of a system with time-independent linear dynamics and Gaussian measurement noise to evaluate schemes for improving effective measurement efficiency.

Key Contributions and Results

Error-Resistant Clifford Gates and Modified Decoding:
The paper demonstrates that GKP Clifford gates generally spread displacement errors, reducing gate fidelity. The authors propose a modified decoding patch strategy where the error correction is performed over a deformed lattice patch $P' = S(\bar{A})V_L$ (where $S(\bar{A})$ is the symplectic matrix of the gate and $V_L$ is the Voronoi cell).
- Ideal Case: In the absence of gate noise, this modification exactly counteracts error spreading, reducing the average gate infidelity by up to two orders of magnitude compared to using the standard Voronoi cell.
- Approximate Case: When accounting for approximate error correction (due to non-ideal codestates), the improvement is reduced to roughly one order of magnitude. The authors provide analytical estimates for the effective distance and degeneracy of these modified patches for both square and hexagonal GKP codes. They find that while the hexagonal code generally outperforms the square code for identity gates, the square code can outperform the hexagonal code for certain gates (like $\bar{C}_{YY}$ ) under specific conditions, though the hexagonal code remains superior for $\bar{C}_{ZZ}$ if a logical Hadamard is applied.
Analytical Estimates for Loss and Dephasing:
The authors derive analytical expressions for the average gate infidelity caused by loss and dephasing on approximate GKP codestates.
- Loss: They show that loss acting on approximate codestates can be modeled as a Gaussian random displacement channel. The variance of this channel depends on both the squeezing parameter ( $\Delta$ ) and the loss rate ( $\gamma$ ), capturing a trade-off where highly squeezed states (small $\Delta$ ) are more susceptible to loss due to higher average photon numbers.
- Dephasing: Unlike loss, dephasing does not contribute to infidelity at first order for small noise levels. However, once dephasing exceeds a critical threshold ( $\sigma_d^*$ ), it has an exponential impact on infidelity. The authors provide formulas for the critical dephasing threshold and the optimal squeezing parameter to minimize infidelity under combined loss and dephasing.
- Kerr Nonlinearity: The paper briefly analyzes Kerr nonlinearities, noting that higher-quality codestates are more severely affected, introducing a similar trade-off to loss.
Improving Logical Read-Out Efficiency:
Recognizing that current homodyne detection efficiencies (60–75%) are insufficient for low logical error rates, the authors analyze two schemes to improve effective measurement efficiency ( $\eta_{eff}$ ):
1. Single-Mode Squeezing: Applying a squeezing operator before measurement. This can boost effective efficiency (e.g., from 0.7 to 0.85) but requires releasing the GKP mode directly into the measurement chain.
2. Ancilla Coupling: Coupling the high-Q GKP mode to a low-Q read-out ancilla via a Hamiltonian $H = -gq_1p_2$ . By performing homodyne detection on the ancilla, the system behaves as an effective measurement of the GKP mode with enhanced efficiency.
- Results: Using quantum trajectories, the authors show that with realistic coupling strengths ( $g \approx 10 \times 2\pi$ MHz) and measurement times ( $\approx 0.45$ –$0.63$ $\mu$ s), effective efficiencies of $\eta_{eff} \approx 0.85$ to $0.92$ can be achieved. This allows for logical measurement errors of 1% and 0.1%, respectively, even with physical efficiencies as low as 0.75.

Significance
The paper claims to bridge the gap between theoretical GKP models and experimental realities in superconducting circuits. By providing a unified framework that incorporates realistic noise channels, gate-induced error spreading, and measurement inefficiencies, the work offers concrete tools for optimizing GKP-based quantum computing. Specifically, the proposed modified decoding patches significantly reduce the resource overhead (squeezing requirements) needed to achieve target gate fidelities, and the read-out schemes demonstrate a pathway to overcoming the limitations of current microwave homodyne detection. The authors emphasize that their analytical approximations agree well with numerical simulations, validating the use of these simplified models for designing future fault-tolerant architectures.

More like this