Bosonic quantum error correction codes encode logical qubits in the Hilbert space of one or multiple harmonic oscillators. A prominent class of bosonic codes is that of Gottesman-Kitaev-Preskill (GKP) codes of which implementations have been demonstrated with trapped ions and microwave cavities. In this paper, we investigate theoretically the preparation and error correction of a GKP qubit in a vibrational mode of a neutral atom stored in an optical dipole trap. This platform has recently shown remarkable progress in simultaneously controlling the motional and electronic degrees of freedom of trapped atoms. The protocols we develop make use of motional states and, additionally, internal electronic states of the trapped atom to serve as an ancilla qubit. We compare optical tweezer arrays and optical lattices and find that the latter provide more flexible control over the confinement in the out-of-plane direction, which can be utilized to optimize the conditions for the implementation of GKP codes. Concretely, the different frequency scales that the harmonic oscillators in the axial and radial lattice directions exhibit and a small oscillator anharmonicity prove to be beneficial for robust encodings of GKP states. Finally, we underpin the experimental feasibility of the proposed protocols by numerically simulating the preparation of GKP qubits in an optical lattice with realistic parameters.
Bosonic quantum error correction codes encode logical qubits in the Hilbert space of one or multiple harmonic oscillators. A prominent class of bosonic codes is that of Gottesman-Kitaev-Preskill (GKP) codes of which implementations have been demonstrated with trapped ions and microwave cavities. In this paper, we investigate theoretically the preparation and error correction of a GKP qubit in a vibrational mode of a neutral atom stored in an optical dipole trap. This platform has recently shown remarkable progress in simultaneously controlling the motional and electronic degrees of freedom of trapped atoms. The protocols we develop make use of motional states and, additionally, internal electronic states of the trapped atom to serve as an ancilla qubit. We compare optical tweezer arrays and optical lattices and find that the latter provide more flexible control over the confinement in the out-of-plane direction, which can be utilized to optimize the conditions for the implementation of GKP codes. Concretely, the different frequency scales that the harmonic oscillators in the axial and radial lattice directions exhibit and a small oscillator anharmonicity prove to be beneficial for robust encodings of GKP states. Finally, we underpin the experimental feasibility of the proposed protocols by numerically simulating the preparation of GKP qubits in an optical lattice with realistic parameters.
Straightforward logical operations contrasting with complex state preparation are the hallmarks of the bosonic encoding proposed by Gottesman, Kitaev, and Preskill (GKP). The recently reported generation and error correction of GKP qubits in trapped ions and superconducting circuits thus holds great promise for the future of quantum computing architectures based on such encoded qubits. However, these experiments rely on error-syndrome detection via an auxiliary physical qubit, whose noise may propagate and corrupt the encoded GKP qubit. We propose a simple module composed of two oscillators and a physical qubit, operated with two experimentally accessible quantum gates and elementary feedback controls to implement an error-corrected GKP qubit protected from such propagating errors. In the idealized setting of periodic GKP states, we develop efficient numerical methods to optimize our protocol parameters and show that errors of the encoded qubit stemming from flips of the physical qubit and diffusion of the oscillators state in phase space may be exponentially suppressed as the noise strength over individual operations is decreased. Our approach circumvents the main roadblock towards fault-tolerant quantum computation with GKP qubits.
A universal, large-scale quantum computer would be a powerful tool with applications of high value to mankind. For example, such a computer could significantly speed up the search for new medications or materials. However, the error rates of current qubit designs are simply too large to enable interesting computations. Therefore, both error correction and improved designs of qubits are needed. In 2001, Gottesman, Kitaev and Preskill proposed an encoding (GKP code) where a qubit is stored in a harmonic oscillator — a system that can be controlled and manufactured with high precision, and therefore have comparatively high coherence times. Moreover, the code offers good protection against losses, a simple gate set, and error correction circuits that are comparatively easy to implement. The drawback is that encoding a qubit into a GKP code state is a challenging task. In this thesis, we develop efficient schemes to encode a GKP qubit. Bosonic codes, where a qubit is stored in an oscillator, and in particular the GKP code are still relatively unknown. Therefore, we will start the thesis with an overview of the field, and provide the reader with the tools to analyze a GKP code, as these are quite different from standard error correcting codes. A tool which is important to understand, and that describes a protocol that encodes a GKP qubit is the so-called phase estimation algorithm. This algorithm allows to measure the eigenvalue of any unitary operation, and is one of the cornerstones of quantum information. We will show how phase estimation can be applied to encode a GKP qubit, and what the requirements for an experiment attempting to do so are. A major advantage of the GKP code over other encodings is that it can tolerate significant photon loss before the encoded information is lost. In addition, states that are closely related to the GKP qubit can be used to violate Bell’s inequalities (i. e. prove the presence of entanglement), even in the presence of large noise. Both these applications make the code particularly interesting in the optical regime, where error correction usually cannot be done while the signal is travelling. In this thesis, we will analyze an encoding protocol originally proposed by H.M. Vasconcelos, L. Sanz, and S. Glancy, Optics Letters 35, 3261 (2010) that relied on post-selection, and show that any output state can be used as a GKP code state with a simple change of frame, providing an exponential speedup. In 2019, two separate experiments generated a GKP code state for the first time: C.Flühmann et al., Nature 566, 513 (2019) realized a GKP qubit in the motional mode of a trapped ion, while P. Campagne-Ibarcq et al., Nature 584, 368 (2020) realized it with a transmon qubit coupled to a microwave cavity. However, both these experiments employ phase estimation, which is slow because it requires many measurements in sequence. We propose a circuit that allows a single-shot measurement of the GKP stabilizers, and analyze the performance of such a measurement as well as the impact of noise.
To be useful, quantum computers will be required to successfully correct errors occurring at the hardware level. Bosonic codes provide a hardware-efficient option for error correction, but fault-tolerance further requires that the available gate interactions be compatible with the code. A promising bosonic code is the Gottesman-Kitaev-Preskill (GKP) code, for which a linear beamsplitter-like coupling between two bosonic modes is fault-tolerant, making this a key primitive for building larger systems. Here, using two motional modes of a trapped ion, we demonstrate the generation of entangled states of GKP qubits by interfering two qunaught states, which have a grid structure but carry no logical information, on a beamsplitter. We generate all four Bell states with an average fidelity of 69%, and subsequently demonstrate an extension of the entangled state lifetime through the use of quantum error correction. These results complete the set of Gaussian operations required for quantum computing with GKP codes and enable explorations of multi-mode bosonic encodings as well as fundamental tests of information channels.
We present a universal logical gate set for finite-energy Gottesman-Kitaev-Preskill (GKP) qubits encoded in the harmonic motional states of trapped neutral atoms. Internal electronic states serve as ancilla for implementing state-dependent conditional displacements in phase space, enabling arbitrary single-qubit phase gates. A Transient Rydberg excitation-mediated atomic dipole-dipole interaction enables a controlled-Z gate without invoking the blockade mechanism. The gates operate under magic-trapping conditions, allowing continuous trapping throughout the gate sequence, thereby preserving the motional encoding without intermediate measurement or feedforward. We assess the experimental feasibility of the proposed protocols using neutral 88Sr atoms and obtain average gate fidelities of 0.986 for single-qubit phase gates and 0.985 for the controlled-Z gate at a finite squeezing parameter of Δ=0.25 (12 dB of squeezing), increasing to 0.997 and 0.995 respectively as Δ reaches 0.1. These results provide a route toward universal, measurement-free logical control of motional GKP qubits in neutral-atom architectures.
State-of-the-art quantum computers are not yet capable of executing large-scale fault-tolerant quantum algorithms. However, experiments in recent years have demonstrated remarkable noisy intermediate-scale and early fault-tolerant quantum computations. To use the existing yet limited hardware most effectively, these protocols are typically tailored to a specific device or hardware platform that comes with certain capabilities but also with individual limitations. This thesis studies quantum protocols suited for near-term atomic quantum processors, i.e., neutral atoms in optical tweezer arrays and trapped ions. One challenge these two platforms have in common is the difficulty of performing mid-circuit measurements. On the other hand, both platforms offer advantages in qubit connectivity and scalability, both of which can be achieved by shuttling atoms or ions. Furthermore, neutral-atom and trapped-ion setups provide native support for multiqubit gates. The protocols explored in this thesis focus on quantum error correction, but also touch on quantum simulation. Concretely, we first investigate quantum autoencoders as a tool for measurement-free quantum error correction in a quantum memory. We propose an experimental implementation of this quantum machine-learning model on a shuttling-based trapped-ion quantum processor to demonstrate that quantum autoencoders can independently identify optimal strategies that combat hardware-specific noise. Secondly, we develop a software toolbox to optimize two- and multiqubit Rydberg gates utilized in neutral-atom quantum processors. We show that multiqubit gates can be useful for achieving good performance and efficient implementations of Floquet and measurement-free fault-tolerant quantum error correction with neutral atoms. Thirdly, we study the feasibility of neutral atoms as a platform for bosonic quantum error correction. We devise blueprints for encoding Gottesman-Kitaev-Preskill qubits in selected motional modes of neutral atoms confined in optical dipole traps and compare the suitability of optical tweezers and optical lattices for this purpose. Finally, we investigate a novel dissipative variational quantum eigen solver that enables the study of ground-state properties of the pure Z-2 lattice gauge theory. We demonstrate that the scheme outperforms established unitary approaches in the presence of circuit-level noise and can be implemented in available neutral-atom quantum processors. Overall, the schemes developed in this work could serve as building blocks for executing elaborate protocols on state-of-the-art atomic quantum processors. Therefore, the results of this thesis contribute to bridging the gap between the theoretical development of protocols for quantum error correction or quantum simulation and their experimental realization.
"Quantum error correction protects fragile quantum information by encoding it redundantly. The Gottesman-Kitaev-Preskill (GKP) code takes an elegant approach: it encodes a qubit in the infinite-dimensional Hilbert space of a single bosonic mode — a microwave cavity field — using a lattice structure in phase space. This thesis develops the theory of the GKP code along a narration from idealization to implementation. We begin with the ideal code's mathematical structure, then confront the two physical constraints: finite energy, which forces normalizable approximations with a squeezing parameter, and photon loss, the dominant noise channel in superconducting circuit QED hardware. We analyze how loss distorts the phase-space lattice and develop the recovery operations — both operational error correction and the near-optimal Petz recovery map — that attempt to restore it. Finally, we present energy-scaled zero-noise extrapolation (ES-ZNE), an error mitigation technique that exploits the GKP code's relationship between photon number and code performance to extrapolate noiseless expectation values from a ladder of finite-energy measurements. "