Courtesy: Xanadu
Fault-tolerant quantum computing remains one of the most significant challenges in the field, largely due to the fragility of qubits and their susceptibility to noise. Among the many approaches being explored, bosonic qubits encoded in photonic systems have emerged as a promising route—particularly through the use of Gottesman–Kitaev–Preskill (GKP) states.
Qubits in Phase Space
Unlike conventional qubits that rely on discrete two-level systems, bosonic qubits exploit the continuous nature of light. The quantum state of light can be represented in phase space, an infinite mathematical plane where each point corresponds to a possible configuration of the electromagnetic field.
Gaussian states—such as coherent states produced by lasers or squeezed states generated using photonic circuits—appear as smooth distributions in phase space. By carefully manipulating these states, it is possible to encode logical qubits directly into the continuous variables of light.
The key challenge, however, lies in protecting these encoded qubits from noise, which typically manifests as small displacements in phase space.
Grid States and Continuous Error Correction
GKP states provide an elegant solution to this problem. They encode logical 0s and 1s as periodic lattice patterns in phase space, analogous to placing pieces on alternating squares of an infinite chessboard. Because of this structured layout, small displacement errors can be detected and corrected by identifying which region—or “square”—a state occupies.
This form of continuous-variable error correction is particularly powerful. Not only can it correct small shifts in phase space, but, when combined across multiple modes, it can also help mitigate larger errors and even photon loss—one of the most severe challenges in photonic quantum systems.
Gates, Measurements, and Universality
Beyond error correction, GKP states support the full set of operations required for quantum computation. Clifford gates, which form the backbone of many quantum algorithms, can be implemented using standard Gaussian optical operations such as displacements and squeezers.
Measurement of GKP qubits is also experimentally accessible. By performing homodyne detection, researchers can determine the phase-space quadrature of a state and infer its logical value based on its position within the lattice.
To achieve universal quantum computation, non-Clifford operations are required. These can be enabled by preparing special superpositions of GKP states, often referred to as magic states, completing the toolkit for fault-tolerant quantum processing.
The Reality of Imperfect GKP States
Ideal GKP states require infinite energy and are therefore unphysical. In practice, laboratories can only produce approximate GKP states, consisting of a finite number of broadened lattice peaks. While these imperfections limit the amount of error that can be corrected by a single state, approximate GKP states remain highly valuable—especially for near-term NISQ (Noisy Intermediate-Scale Quantum) devices.
Recent research has focused on quantifying how these imperfections affect encoding quality, gate performance, and error correction, offering important guidance for scalable system design.
Creating GKP States in Photonic Hardware
Photonic platforms provide a natural environment for generating and manipulating bosonic qubits. Using squeezers, displacers, interferometers, and homodyne detectors, researchers can prepare and control Gaussian states with high precision.
To move beyond Gaussian states and generate GKP states, photon-number-resolving detectors (PNRs) play a critical role. These detectors probabilistically project Gaussian light into non-Gaussian states by selectively measuring photon numbers in specific modes.
A typical preparation scheme involves generating a multi-mode Gaussian state, entangling it through an interferometer, and then measuring all but one mode with PNRs. With the right configuration—and some probability of success—the remaining mode collapses into an approximate GKP state.
The main trade-off lies between state quality and generation probability, an optimization challenge that is increasingly being addressed using machine-learning techniques.
Looking Ahead
As photonic losses decrease and non-Gaussian operations improve, the efficiency and fidelity of GKP state preparation are expected to rise significantly. With GKP encoding offering a unified framework for error correction, gate implementation, and measurement, bosonic qubits stand out as a compelling path toward scalable, fault-tolerant quantum computing.
