Hyperbolic Cluster States for Fault-Tolerant Measurement-Based Quantum Computing

Fault-tolerant measurement-based quantum computing (MBQC) provides a compelling framework for fault-tolerant quantum computation, in which quantum information is processed through single-qubit measurements on a three-dimensional entangled resource known as cluster state. To date, this resource has been predominantly studied on Euclidean lattices, most notably in the Raussendorf-Harrington-Goyal (RHG) construction, which underlies topological fault tolerance in MBQC. In this work, we introduce the hyperbolic cluster state, a generalization of the three-dimensional cluster state to negatively curved geometries, obtained via the foliation of periodic hyperbolic lattices. We present an explicit construction of hyperbolic cluster states and investigate their fault-tolerant properties under a realistic circuit-level depolarizing noise model. Using large-scale numerical simulations, we perform memory experiments to characterize their logical error rates and decoding performance. Our results demonstrate that hyperbolic cluster states exhibit a fault-tolerance threshold comparable to that of the Euclidean RHG cluster state, while simultaneously supporting a constant encoding rate in the thermodynamic limit. This represents a substantial improvement in qubit overhead relative to conventional cluster-state constructions. These findings establish hyperbolic geometry as a powerful and experimentally relevant resource for scalable, fault-tolerant MBQC and open new avenues for leveraging negative curvature in quantum information processing.

💡 Research Summary

The paper introduces a novel three‑dimensional (3D) cluster state built on hyperbolic (negatively curved) lattices, extending the well‑known Raussendorf‑Harrington‑Goyal (RHG) construction that is based on Euclidean cubic lattices. The authors first review the mathematics of regular {p,q} hyperbolic tilings, emphasizing that the curvature condition 1/p + 1/q < 1/2 admits an infinite family of lattices, each of which can be represented in the Poincaré disk model. By imposing periodic boundary conditions (PBC) on a finite patch of such a tiling, they obtain a closed surface of genus g ≥ 2 (Euler characteristic χ < 0). This high‑genus topology is the source of a constant encoding rate: for a lattice with E edges (physical qubits) the number of logical qubits is k = 2g, yielding k/E → 1 − 2/p − 2/q as the system size grows. In contrast, Euclidean surface codes have k/E → 0, so the hyperbolic construction promises a dramatic reduction in qubit overhead.

The hyperbolic CSS code is defined by placing qubits on edges, Z‑type stabilizers on faces, and X‑type stabilizers on vertices. Logical operators correspond to non‑trivial cycles and co‑cycles on the underlying graph. Although the code distance scales only logarithmically with the number of qubits (d = O(log n)), the authors argue that the trade‑off between distance and encoding rate can be tuned by choosing larger p or q, which increase stabilizer weight and thus the encoding rate at the expense of a modestly lower threshold.

To turn this code into a measurement‑based quantum computing (MBQC) resource, the authors employ the foliation technique. Starting from a 2D hyperbolic CSS code, they create a “primal” layer by preparing data qubits in |+⟩ and adding an ancilla qubit for each Z‑type stabilizer; CZ gates entangle each ancilla with the data qubits it checks. Measuring the ancilla in the X basis projects the data qubits onto the eigenstate of the corresponding plaquette operator, thereby extracting the Z‑type syndrome. A dual layer is built analogously for X‑type checks, and alternating primal and dual layers are coupled via CZ gates between corresponding data qubits, yielding a full 3D foliated cluster state. In this picture, the discrete layer index plays the role of a time coordinate, and intermediate measurement outcomes provide syndrome information for error correction.

The authors evaluate fault‑tolerance using large‑scale Monte‑Carlo simulations under a realistic circuit‑level depolarizing noise model (including gate, measurement, and preparation errors). They perform memory experiments, initializing logical information, allowing it to evolve through several foliated layers, and finally measuring it out. Logical error rates are extracted as a function of physical error probability. The hyperbolic cluster state exhibits a threshold p_th ≈ 0.62 %, essentially comparable to that of the Euclidean RHG cluster state. Crucially, for the same physical qubit budget the hyperbolic construction supports 5–10 times more logical qubits, confirming the constant‑rate advantage.

Decoding is performed with a minimum‑weight perfect‑matching (MWPM) algorithm adapted to the hyperbolic graph’s non‑planar geometry. The authors also describe how they generate the underlying lattices algorithmically: using GAP and the LINS package they enumerate finite‑index normal subgroups of the relevant Fuchsian groups, enabling systematic construction of periodic hyperbolic lattices for arbitrary {p,q} parameters.

From an experimental perspective, the paper notes that hyperbolic connectivity can be realized on platforms that allow long‑range couplings, such as superconducting resonator arrays, trapped‑ion chains, or photonic integrated circuits. Because the physical layout remains two‑dimensional, existing fabrication techniques can be leveraged, while the logical connectivity is effectively hyperbolic due to engineered couplings.

In summary, the work demonstrates that hyperbolic cluster states retain the high fault‑tolerance threshold of traditional RHG states while providing a constant logical‑to‑physical qubit ratio, thereby reducing the overhead required for scalable fault‑tolerant quantum computation. This establishes negative curvature as a powerful resource‑engineering tool for MBQC and opens avenues for “curvature‑engineered” quantum architectures, including exploration of higher‑p,q tilings, heterogeneous hyperbolic‑Euclidean hybrids, and hardware‑specific optimization of the foliation schedule.