Structural Conditions for Native CCZ Magic-State Fountains in qLDPC Codes

Quantum low-density parity-check (qLDPC) codes promise constant-rate, linear-distance families with bounded-weight checks, and recent work has realized transversal or constant-depth non-Clifford gates on various (often non-LDPC) codes. However, no explicit \emph{qubit} qLDPC family is known that simultaneously has constant rate, linear distance, bounded stabilizer weight, and a native \emph{magic-state fountain} that prepares many non-Clifford resource states in constant depth. We take a structural approach and identify coding-theoretic conditions under which a CSS qLDPC family necessarily supports a constant-depth $\CCZ$ magic-state fountain. The key ingredients are: (i) an algebraic notion of \emph{magic-friendly triples} of $X$-type logical operators, defined by pairwise orthogonality and a triple-overlap form controlling diagonal $\CCZ$ phases, and (ii) a 3-uniform hypergraph model of physical $\CCZ$ circuits combined with a packing lemma that turns large collections of such triples with bounded overlaps into bounded-degree hypergraphs. Our main theorem shows that if a CSS code family on $n$ qubits admits $Ω(n^{1+γ})$ magic-friendly triples whose supports have bounded per-qubit participation, then there exists a constant-depth circuit of physical $\CCZ$ gates implementing $Ω(n^γ)$ logical $\CCZ$ gates in parallel while preserving distance up to a constant factor. For asymptotically good qLDPC families such as quantum Tanner codes, this reduces the existence of a native $\CCZ$ magic-state fountain to a concrete combinatorial problem about counting and distributing magic-friendly triples in the logical $X$ space.

💡 Research Summary

This paper tackles a central challenge in fault‑tolerant quantum computing: finding a family of qubit‑level quantum low‑density parity‑check (qLDPC) codes that simultaneously enjoys constant rate, linear distance, bounded‑weight stabilizers, and a “magic‑state fountain” that can produce many non‑Clifford resource states (specifically CCZ states) in constant depth. While previous works have demonstrated transversal or constant‑depth non‑Clifford gates on various codes, none of the known qubit qLDPC families satisfy all four desiderata at once.

The authors adopt a structural rather than constructive approach. They identify algebraic and combinatorial conditions on the logical X operators of any CSS qLDPC code that guarantee the existence of a constant‑depth physical CCZ circuit implementing a large number of logical CCZ gates in parallel, while preserving the code distance up to a constant factor.

Key definitions

1. Magic‑friendly triple – a triple ((x,y,z)) of logical X operators (vectors in the dual of the Z‑stabilizer space) that satisfies three properties: (i) the three vectors are linearly independent in the logical X space, (ii) each pair is orthogonal over (\mathbb{F}_2) (pairwise inner product zero), and (iii) the triple overlap (\tau(x,y,z)=\sum_i x_i y_i z_i) is odd. When such a triple is realized physically by applying a CCZ gate on each qubit index (i) simultaneously, the overall phase contributed to the computational basis state is ((-1)^{\tau(x,y,z)}). Hence the triple implements a logical CCZ on the three logical qubits associated with (x,y,z), up to known Clifford corrections.

2. 3‑uniform hypergraph model – a set of physical CCZ gates is represented as a hypergraph (H=(V,E)) where vertices are physical qubits and each hyperedge (e\in E) (size three) corresponds to one CCZ gate. The degree (\deg_H(v)) counts how many CCZ gates involve qubit (v). If the maximum degree (\Delta) is bounded by a constant, a classic edge‑coloring argument shows that the hypergraph can be colored with at most (3\Delta+1) colors, each color representing a layer of mutually disjoint CCZ gates. Consequently, the entire circuit depth is bounded by a constant independent of the code size.

Packing lemma – The paper proves that if we have a collection ({S_t}) of support sets (each being the union of the supports of a magic‑friendly triple) such that (a) each (|S_t|) scales linearly with the total number of qubits (n) (i.e., (a n \le |S_t| \le b n) for constants (a,b)), and (b) each physical qubit belongs to at most a constant (M) of these support sets, then a large sub‑collection can be extracted where the supports are pairwise disjoint. Specifically, at least (|\mathcal{S}|/(M b)) triples survive the greedy selection. This guarantees that many triples can be used simultaneously without over‑loading any qubit.

Main theorem – Combining the hypergraph coloring and packing results, the authors show: if a CSS qLDPC family ({Q_n}) on (n) qubits possesses (i) (\Omega(n^{1+\gamma})) magic‑friendly triples for some (\gamma>0), (ii) each triple’s support size is (\Theta(n)), and (iii) each qubit participates in at most a constant number (M) of triples, then there exists a sub‑collection of (\Omega(n^{\beta})) triples (for some (\beta>0)) whose associated hypergraph has bounded degree (\Delta). Consequently, a constant‑depth physical CCZ circuit implements (\Omega(n^{\beta})) logical CCZ gates in parallel, up to known Clifford frames. Moreover, if the original code distance scales linearly ((d_n = \Theta(n))), the distance after the circuit remains linear up to a constant factor, as proved via a light‑cone argument for constant‑depth local circuits.

Examples – The paper provides two toy examples: a 4‑qubit trivial CSS code that admits a single magic‑friendly triple, and a doubled version yielding two disjoint triples with degree 1 hypergraph, illustrating the construction of a depth‑2 CCZ circuit. While these examples are not asymptotically good, they make the abstract combinatorics concrete.

Implications and contributions

1. The work isolates a precise combinatorial property—abundance of low‑overlap magic‑friendly triples—that any qLDPC code must exhibit to support a native CCZ magic‑state fountain. This unifies earlier constructions based on tri‑orthogonal matrices, cup‑product structures, and product‑code techniques under a single algebraic condition on logical X operators.
2. It reframes the search for non‑Clifford gates on qLDPC codes as an explicit problem in high‑dimensional expander combinatorics: prove that a given asymptotically good qLDPC family (e.g., quantum Tanner codes) contains (\Omega(k)) magic‑friendly triples with constant per‑qubit participation. Success immediately yields a linear‑size magic‑state fountain with LDPC checks and linear distance.
3. The paper supplies a reusable “recipe”: pack logical operators → construct a bounded‑degree hypergraph → color it → obtain a constant‑depth diagonal circuit with provable distance preservation. The authors anticipate that the same pattern can be adapted to multi‑controlled‑Z factories and other product‑based qLDPC families.

Open problems – The authors stress that they do not prove the existence of the required triples for any concrete good qLDPC family; establishing such bounds remains an open combinatorial challenge. Demonstrating that quantum Tanner codes or the product‑code families of Refs.