Fast magic state preparation by gauging higher-form transversal gates in parallel

Magic states are a foundational resource for universal quantum computation. To survive in a realistic noisy environment, magic states must be prepared fault-tolerantly and protected by a quantum error-correcting code. The recent discovery of highly efficient quantum low-density parity-check codes, together with efficient logic gates, lays the groundwork for low-overhead fault-tolerant quantum computation. This motivates the search for fast and parallel protocols for logical magic state preparation to enable universal quantum computation. Here, we introduce a fast code surgery procedure that performs a fault-tolerant measurement of many transversal logic gates in parallel. This is achieved by performing a generalized gauging measurement on a quantum code that supports a higher-form transversal gate. The time overhead of our procedure is constant, and the qubit overhead is linear. The procedure inherits fault-tolerance properties from the base code and the structure of the higher-form transversal gate. When applied to codes that support higher-form Clifford gates our procedure achieves fast and fault-tolerant preparation of many magic states in parallel. This motivates the search for good quantum low-density parity-check codes that support higher-form Clifford gates.

💡 Research Summary

The paper addresses a central bottleneck in fault‑tolerant quantum computing: the efficient, parallel preparation of many magic states, which are required to implement non‑Clifford gates such as T and CCZ. Recent advances in quantum low‑density parity‑check (qLDPC) codes have shown that one can obtain high distance and low‑weight stabilizer checks simultaneously, opening the door to low‑overhead fault‑tolerant architectures. However, existing magic‑state preparation techniques either rely on sequential gate teleportation, costly distillation, or the existence of transversal non‑Clifford gates, all of which incur substantial space‑time overhead.

The authors propose a fundamentally different approach based on higher‑form transversal gates. In the language of topological quantum codes, a conventional 0‑form symmetry corresponds to a global operator acting uniformly on the whole system. By contrast, an h‑form symmetry acts on codimension‑h submanifolds; mathematically it is captured by the kernel of a coboundary map (\delta_{h+1}) in a chain complex that describes the qLDPC code. An h‑form transversal gate is defined as a product of on‑site unitaries (U_s) indexed by the sites of the complex, with the pattern of activation given by a chain (c\in\ker\delta_{h+1}). The authors focus on 1‑form Clifford gates, where the on‑site unitaries are Pauli‑type (e.g., (X), controlled‑(Z), or (\sqrt{-i}X,S)). These gates are “strongly transversal” when all (U_s) are identical, but the definition is relaxed to allow site‑dependent unitaries as long as they form a linear (non‑projective) representation of (\mathbb{Z}_2) and satisfy sparsity constraints required by qLDPC codes.

The core technical contribution is a higher‑form gauging measurement protocol (Algorithm 1). The procedure works as follows:

1. Auxiliary qubits are introduced on the (h + 1)-cells (hyperedges) of the chain complex and initialized in (|0\rangle).
2. For each h‑cell (vertex in the 1‑form case) a joint measurement of the operator (A_v = U_v \prod_{e\in\delta_{h+1}v} X_e) is performed. The outcome (\epsilon = \pm 1) updates the logical eigenvalue (\sigma_i) associated with each generator (\ell_i) of the h‑th cohomology group.
3. After all vertex measurements, each auxiliary qubit is measured in the Z basis, yielding a binary vector (x). Solving the linear system (\delta_{h+1} y = x) (via Gaussian elimination) produces a correction chain (y).
4. Finally, the physical operator (U(y)) is applied, completing the gauging of the higher‑form symmetry.

All vertex and hyperedge measurements can be performed in parallel, so the circuit depth is a constant independent of the code size. The number of auxiliary qubits scales linearly with the number of (h + 1)-cells, which for 1‑form gates is proportional to the number of edges in the underlying hypergraph, i.e., linear in the total number of physical qubits. Because the underlying qLDPC code has growing distance, any constant‑size error can be locally corrected, and the gauging measurement inherits the fault‑tolerance of the base code. The authors emphasize that the higher‑form symmetry must be anomaly‑free (i.e., gaugeable); this is ensured by requiring the on‑site representation to be a genuine linear representation rather than a projective one, and by demanding that the homology/cohomology groups of the symmetry have growing distance.

Applying this framework to 1‑form Clifford gates yields concrete magic‑state preparation protocols:

• 1‑form CZ gates: When a CSS code supports a transversal CZ (or more generally a 0‑form non‑Clifford gate), the associated 1‑form CZ gate can be measured using the gauging protocol. Measuring a strongly transversal 1‑form CZ on a code prepared in the X basis produces CCZ magic states on the logical qubits. Importantly, the existence of a 0‑form CZ is not required; any pair of CSS codes whose X‑type check matrices satisfy a simple element‑wise product condition automatically defines a 1‑form CZ symmetry.

• 1‑form (\sqrt{-i}X,S) gates: Analogously, measuring a 1‑form (\sqrt{-i}X,S) on a code initialized in a Pauli basis yields T (or (T^\dagger)) magic states. The condition for the symmetry to preserve the code space reduces to an element‑wise product constraint similar to the CZ case, together with a parity condition on overlapping support regions (e.g., multiples of four sites or equal numbers of (X S) and (X S^\dagger) actions).

Both protocols generate many magic states in a single constant‑time round, a dramatic improvement over sequential distillation or teleportation schemes. Because the measurement remains within the stabilizer formalism, no additional non‑Clifford resources are needed during the preparation phase.

The paper also discusses the implications for code design. Prior work has focused on finding qLDPC codes that admit transversal non‑Clifford gates (e.g., CCZ or T) or that support magic‑state distillation with low overhead. The present work shows that supporting only 1‑form Clifford gates is sufficient for fast magic‑state generation, thereby widening the class of candidate codes. The authors call for the discovery of qLDPC families with single‑shot Pauli state preparation and 1‑form Clifford symmetries, which would combine the best features of low‑density checks, high distance, and constant‑time magic‑state supply.

In summary, the authors introduce a novel, constant‑depth, linear‑overhead protocol for parallel magic‑state preparation based on gauging higher‑form transversal gates. The method leverages the algebraic structure of qLDPC codes, extends the notion of transversal gates to higher‑form symmetries, and provides explicit constructions for 1‑form CZ and (\sqrt{-i}X,S) gates. By doing so, it opens a new pathway toward scalable, low‑overhead fault‑tolerant quantum computation, shifting the focus from finding transversal non‑Clifford gates to engineering codes with suitable higher‑form Clifford symmetries. Future work will need to identify concrete code families, integrate efficient decoders, and experimentally validate the constant‑time gauging measurement in realistic quantum hardware.