#gkpqubits

Gottesman Kitaev Preskill

Sydney Leads: Hardware-Efficient GKP Qubits Create Quantum Logic Gates

Sydney University researchers and Q-CTRL scientists proved a universal quantum logic gate using Gottesman-Kitaev-Preskill (GKP) error-correcting codes, advancing practical, fault-tolerant quantum computing. This groundbreaking Nature Physics experiment shows the entanglement of two logical qubits within the intrinsic vibrations of a confined ytterbium atom, reducing the amount of physical qubits needed for such operations.

New theories on GKP Clifford gates and state read-out in superconducting circuits are opening the door to hardware-efficient quantum computation.

GKP codes have long been touted as a way to build fault-tolerant quantum computers because they can encode logic in a quantum harmonic oscillator's infinite Hilbert space. This allows resilient logical qubits from one physical device.

GKP codes, known as the “Rosetta Stone” of quantum computing, translate quantum systems' continuous oscillations into discrete, digital-like states for mistake detection and repair. GKP Pauli- or Hadamard-eigenstates with Gaussian resources alone can also achieve universal quantum processing.

GKP codes are intriguing in theory but challenging to implement. Approximation states with errors are needed since perfect GKP codes are non-normalizable and impractical. Practical implementation requires detailed quantum system control and error-tolerant, empirically achievable, noise-cancelling methods.

In superconducting circuits, homodyne detection in microwave circuits is a major limitation, with state-of-the-art experiments giving 60% to 75% efficiency. In addition, multi-mode simulations have often used implausible noise models such Gaussian random displacement channels, which fail to accurately depict GKP code performance under loss.

Sydney University Quantum Control Breakthrough

This theoretical promise has been realised by the Sydney Nano Institute's Quantum Control Laboratory at the University of Sydney. They built an entangling logic gate on a trapped ytterbium atom using a laser setup and a Paul trap at ambient temperature. This innovative solution reduces hardware requirements by storing two error-correctable logical qubits in a confined particle's internal quantum vibrations.

Dr. Tingrei Tan, a Sydney Horizon Fellow, said the tests proved that high-quality quantum controls are necessary for many logical qubits. This work lays the framework for hardware-efficient large-scale quantum information processing. Q-CTRL, a University of Sydney spin-off, developed quantum control software that enabled these perfectly calibrated methods. This software produces quantum gates that minimise GKP state distortions during quantum computing while keeping their delicate structure using physics-based modelling.

New Practical Ideas for Superconducting GKP Qubits

A recent study proposes three key concepts to improve GKP qubits, especially in superconducting circuits:

The paper offers executing Clifford circuits without physical single-qubit gates to reduce the number of physical gates and error spread. Software tracks and absorbs single-qubit Clifford gates into two-qubit “generalized controlled gates”. Superconducting circuits may easily build two-qubit gates with a single piece of hardware, distinguishable solely by the phase of a local oscillator in a mixing circuit with three or four waves. Three-wave mixing is better than four-wave mixing because it avoids Kerr, cross-Kerr, and AC Stark shift terms that degrade gate quality.

Modified Decoding Clifford Gates withstand errors: A general Clifford gate fault prevention method uses a modified error correcting mechanism after each gate. Error rectification over a dynamically updated “patch” following gate application can reduce Clifford gates' average gate infidelity by many orders of magnitude to match the identical gate. This change can increase average gate infidelity by roughly two orders of magnitude for square and hexagonal GKP codes when applied to a CZZ gate with GKP squeezing. The “effective distance” and “degeneracy” of the improved decoding patch that accounts for error spreading are calculated using this method. The hexagonal code may perform better than the square code for identity gates, but Clifford gates often reduce the hexagonal code patch's distance, making the modified patch even more significant.

Improved Logical State Read-out: A approach that links each high-Q GKP mode to a low-Q read-out ancilla is presented to overcome homodyne measurements' inefficiencies in superconducting circuits. This improves the logical readout's measurement efficiency. Continuous homodyne detection on the ancilla mode's position quadrature is part of the Hamiltonian-connected technique to the GKP mode. This technique is expected to achieve a 0.1% error rate in 630 nanoseconds, similar to logical reading in transman GKP qubits, with 75% physical efficiency. Strong coupling between the GKP and ancilla modes and GKP mode squeezing reduce logical measurement error. The authors acknowledge that this scheme's performance depends heavily on the quadrature-quadrature coupling and requires a vacuum-initialized low-Q read-out ancilla.

Noise Analysis Tools

In addition to these important insights, the work proposes a novel theoretical method for analytically calculating GKP code loss and dephasing due to typical noise channels. This analytical method justifies and generalises the "twirling approximation," which is used in GKP-qubit code concatenation studies to represent the non-unitary envelope operator as a Gaussian random displacement channel. The analytical approach can calculate an ideal GKP squeezing level that minimises error for a given loss and/or dephasing. Current experimental loss and dephasing rates suggest an optimal GKP squeezing of 9.3 dB, which matches empirically measured values.

To Scalable Quantum Computing