Single-shot magic state factories for quantum LDPC codes

Detailed Research Summary: Single-Shot Magic State Factories for Quantum LDPC Codes

This research presents a comprehensive, end-to-end protocol for producing and utilizing high-fidelity magic states within a fault-tolerant quantum computation architecture based on Quantum Low-Density Parity-Check (qLDPC) codes. The central innovation is the development of a space-time efficient magic state factory that achieves significantly lower space-time cost compared to existing methods, while maintaining exceptionally low logical error rates.

Core Concept and Architecture

The factory is designed to separate the complex tasks of magic state generation, storage, and processing by assigning each function to a code optimized for that specific role. The protocol leverages the complementary strengths of two distinct families of codes: Tricycle codes and four-dimensional balanced product quadcycle codes.

1. Tricycle Codes: These are utilized for their ability to support transversal non-Clifford gates. They can generate multiple logical magic states in parallel. However, their strongly asymmetric distances impose limitations on fidelity, typically requiring full postselection based on stabilizer measurements to achieve low logical error rates.

2. Quadcycle Codes: These codes offer larger and more balanced distances, coupled with the capability for single-shot state preparation and error correction for both X and Z basis stabilizers. Crucially, they lack transversal non-Clifford gates but excel at storing the resulting magic states with sufficient distance to maintain fidelity through subsequent error correction rounds.

The Key Mechanism: Space-Time Efficient Teleportation

The breakthrough lies in the efficient mechanism that teleports postselected logical magic states from the tricycle codes into the larger quadcycle codes. This is achieved by embedding every tricycle code within a corresponding quadcycle code, enabling a depth-1 transversal homomorphic CNOT gate between the two families.

• Teleportation Advantage: The measurement outcomes from this transversal teleportation circuit provide necessary information for efficient and accurate postselection, which suppresses infidelity by several orders of magnitude.

• Storage Fidelity: The resulting postselected magic state is natively stored in the quadcycle code, which possesses the necessary distance to maintain fidelity across subsequent error correction rounds.

Protocol Stages

The factory operates in a structured three-stage protocol:

1. State Preparation (Tricycle Blocks): Three tricycle blocks are prepared in the + state and subjected to a depth-1 transversal CCZ circuit, resulting in the logical hypergraph state.

2. Teleportation and Postselection: Simultaneously, three paired quadcycle blocks are prepared in the 0 state. A depth-1 transversal CNOT then teleports the logical state from the tricycle blocks into the quadcycle blocks. Subsequent transversal measurements of the tricycle blocks yield syndrome information used for postselection.

3. State Extraction and Processing: The resulting logical state is processed to yield two disjoint CCZ states, which are then ready for downstream computation.

Circuit Efficiency and Code Properties

The protocol achieves remarkable efficiency through several code-specific properties:

• Transversality Preservation: A fan-out gadget is employed on top of the tricycle code blocks to convert the depth-2 CCZ circuit into a strictly transversal depth-1 circuit, thereby preserving the overall circuit-level distance of the factory. This is accomplished by pairing each tricycle block with an ancilla patch and entangling them with one transversal layer of CNOTs.

• Shallow Depth: The system achieves constant-depth preparation and postselection due to the single-shot nature of both code families in their relevant bases, keeping the protocol shallow.

• Dimension Jump Implementation: The dimension jump teleportation is realized via a single transversal layer of physical CNOT gates whose control acts on tricycle qubits and target acts on quadcycle qubits.

Implementation Platform: Neutral Atom Arrays

The practical implementation is focused on reconfigurable neutral-atom quantum computers. The product structure of the codes naturally maps onto a regular toric layout, allowing for coordinated syndrome extraction and parallel operation.

• Syndrome Extraction Simplification: The quasi-cyclic and product structure of the lifted product codes facilitates parallel atom moves. Only partial syndrome extraction is required: X checks for the quadcycle code and Z checks for the tricycle codes. This allows simplification by performing CNOTs between all qubit pairs specified by component circulant matrices, minimizing entangling-gate depth to the maximum weight of the measured checks.

• Transversal CCZ Realization: The unit cell layout is specifically designed to allow highly parallel implementation of the transversal CCZ gate, utilizing specific sector arrangements and sequential layers (two layers for a total depth of two) that can be concatenated into a single physical flash, minimizing transport overhead.

Performance Metrics

The factory delivers CCZ states at an error rate as low as 10-9 under a standard circuit-level noise model at physical error rates of 10-3. Crucially, this is achieved with 10–40 times lower space-time cost than state-of-the-art magic state generation methods.

The final implementation on neutral atom arrays shows that the circuit depth for an instance with distance d 3D=3 reaches an effective error exponent of 3.11 plus or minus 0.04, aligning with the theoretical bound. The overall circuit depth is bounded by 3 cubed tau = 27 tau for a specific instance, where tau is the time to traverse the code region diagonally.

Post-Processing and Injection

The protocol concludes with necessary post-processing steps: extracting disjoint CCZ states using homomorphic measurement gadgets (which involve logical Z measurements on ancilla blocks), performing Clifford fix-ups, and injecting the final magic state into another quadcycle block via qLDPC code surgery, leveraging the disjoint logical Z operators.

In summary, this work presents a highly optimized, practical blueprint for generating high-fidelity magic states using qLDPC architectures by ingeniously combining transversal gate capabilities (Tricycle codes) with robust storage and error correction (Quadcycle codes) through an efficient teleportation mechanism.

1. The AI system can perform magic state generation for high-rate qLDPC codes with 10–40 times lower space-time cost than state-of-the-art magic state generation methods by combining transversal non-Clifford gates of recently developed tricycle codes with single-shot four-dimensional balanced product quadcycle codes.

2. The system can utilize an AI search to identify suitable receiving codes, as the authors identify the quadcycle codes by an AI-assisted search and show that they can serve as excellent memories capable of single-shot error correction.

3. The system can produce high-fidelity magic states with logical error rates as low as 10−9 under a standard circuit-level noise model at physical error rate 10−3 with 10–40 times lower space-time cost than state-of-the-art magic state generation methods.

4. The system can be implemented on reconfigurable neutral atom quantum computers, utilizing an end-to-end, practical protocol for producing and utilizing high-fidelity magic states within a high-rate qLDPC architecture, with the lowest space-time cost to date.

5. The AI system can execute constant depth preparation and postselection by using the construction where the fan-out gadget reduces this circuit from depth two to depth one and thereby preserves the circuit-level distance of the factory.

6. The system can achieve a fault distance governed by dX, which is at most d, indicating that the fault distance of the circuit is min(d, dX).

7. The system can perform efficient magic state generation with a space-time cost calculated as cost = raw space-time volume acceptance fraction × number of magic states (6), yielding results roughly an order of magnitude lower than cultivation protocols.

8. The system can be adapted for efficient consumption by utilizing gadgets to extract two disjoint CCZ⟩ states from the hypergraph state det⟩, which is optimal for this state in terms of yield.

Quantum low-density parity-check (qLDPC) codes can substantially reduce the qubit overhead of fault-tolerant quantum computation, but any useful architecture based on such codes also requires an efficient mechanism to implement non-Clifford gates in a fault-tolerant manner. We introduce a space-time-efficient magic state factory for high-rate qLDPC codes by combining the transversal non-Clifford gates of recently developed tricycle codes with single-shot four-dimensional balanced product quadcycle codes. The key ingredient is an efficient "escape" mechanism that teleports postselected logical magic states from tricycle codes to larger distance quadcycle codes. Under a standard circuit-level noise model at physical error rate 10-3, our factories produce CCZ states at logical error rates as low as 10-9 with 10-40 times lower space-time cost than state-of-the-art magic state generation methods. The resulting states are directly encoded in high-rate blocks suitable for computation. We identify the quadcycle codes by an AI-assisted search and show that they can serve as excellent memories capable of single-shot error correction. Finally, we present a concrete implementation of the entire factory with highly efficient atom transport on reconfigurable neutral-atom quantum computers. Together, these results provide an end-to-end, practical protocol for producing and utilizing high-fidelity magic states within a high-rate qLDPC architecture, with the lowest space-time cost to date.