Structural Conditions for Native CCZ Magic-State Fountains in qLDPC Codes
Mohammad Rowshan · 2026
Read Structural Conditions for Native CCZ Magic-State Fountains in qLDPC Codes by Mohammad Rowshan (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-22489.txt.
Introduction
Mohammad Rowshan sets the terms a quantum code must meet to serve as a fountain of magic states. Magic states fuel the hard gates that make quantum machines more than classical ones. His 2026 preprint works with qLDPC codes, which promise steady rates and long reach with light checks.
No known family of this kind yet meets all four marks at once: steady rate, long reach, light checks, and native three-way gates. Rowshan lays out the structural terms that would allow it. Clear terms guide the search for such a family. A find would cut the cost of quantum computing a great deal.
Catalog record
| CREATOR | Mohammad Rowshan |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 27 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2601-22489
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Structural Conditions for Native CCZ Magic-State Fountains in qLDPC Codes Mohammad Rowshan Abstract 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 qubit qLDPC family is known that simultaneously has constant rate, linear distance, bounded stabilizer weight, and a native 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 \mathsf{CCZ} magic-state fountain. The key ingredients are: (i) an algebraic notion of magic-friendly triples of X X -type logical operators, defined by pairwise orthogonality and a triple-overlap form controlling diagonal 𝖢𝖢𝖹 \mathsf{CCZ} phases, and (ii) a 3-uniform hypergraph model of physical 𝖢𝖢𝖹 \mathsf{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 n qubits admits Ω ( n 1 + γ ) \Omega(n^{1+\gamma}) magic-friendly triples whose supports have bounded per-qubit participation, then there exists a constant-depth circuit of physical 𝖢𝖢𝖹 \mathsf{CCZ} gates implementing Ω ( n γ ) \Omega(n^{\gamma}) logical 𝖢𝖢𝖹 \mathsf{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 \mathsf{CCZ} magic-state fountain to a concrete combinatorial problem about counting and distributing magic-friendly triples in the logical X X space. Introduction. Fault-tolerant quantum computation relies on quantum error-correcting codes and the quantum threshold theorem 1 ; 16 ; 18 ; 2 . Surface and color codes 17 ; 4 ; 19 provide local stabilizers compatible with 2D architectures but incur large overheads for magic-state distillation 7 ; 6 ; 10 . Even with optimized factories and compilation strategies 11 ; 25 , the magic-state cost remains a major bottleneck. Quantum LDPC (qLDPC) codes are a promising route to reduce this overhead: they can achieve constant rate and at least polynomial distance with bounded-weight stabilizers 22 ; 20 ; 9 . Architectures and protocols based on high-rate LDPC-like codes are beginning to appear, including high-threshold, low-overhead quantum memories 5 , neutral-atom implementations of high-rate LDPC codes with limited long-range interactions 24 , and homological product codes with fast, parallelizable logical computation 27 . At the same time, locality and limited connectivity impose stringent tradeoffs between rate, distance, and geometric nonlocality 3 ; 26 . In parallel, there has been rapid progress on codes with transversal or locality-preserving non-Clifford gates, motivated by further reducing the cost of magic-state distillation. Asymptotically good quantum codes with transversal \mathsf{CCZ} gates were constructed by Golowich and Guruswami 12 , while Golowich and Lin, and Lin independently, gave qLDPC-type families with transversal non-Clifford gates based on products and sheaf-theoretic constructions 13 ; 21 . He et al. developed a general framework of addressable orthogonality and used it to obtain asymptotically good codes with addressable transversal \mathsf{CCZ} gates 15 . For qLDPC codes specifically, Zhu formulated a topological theory for product-based constructions and demonstrated constant-depth non-Clifford gates and a \mathsf{CCZ} magic-state “fountain” on 3D homological product codes with constant rate and polynomial distance 28 . Despite these advances, we still lack an explicit qubit qLDPC family that simultaneously has (i) bounded stabilizer weight and bounded check degree, (ii) constant rate with linear distance, and (iii) a native constant-depth magic-state fountain that prepares a number of logical \mathsf{CCZ} magic states scaling linearly with block length. The constructions of Refs. 12 ; 15 are asymptotically good but not LDPC; the qLDPC families of Refs. 13 ; 21 relax distance to sublinear scalings; and the homological product codes 28 achieve constant rate and a \Theta(\sqrt{N}) -sized fountain but fall short of linear distance. In this work, we take a complementary, structural point of view. Instead of proposing a new code family, we identify algebraic and combinatorial conditions on the logical X operators of an arbitrary CSS code that are sufficient to support a constant-depth \mathsf{CCZ} magic-state fountain while preserving distance up to a constant factor. Concretely, we introduce the notion of a magic-friendly triple of logical X operators, defined by pairwise orthogonality and an odd triple overlap. We then model physical \mathsf{CCZ} circuits as bounded-degree 3-uniform hypergraphs and prove a structural theorem: if a CSS code family on n qubits has \Omega(n^{1+\gamma}) magic-friendly triples whose supports have bounded per-qubit participation, then there exists a constant-depth circuit of physical \mathsf{CCZ} gates implementing \Omega(n^{\gamma}) logical \mathsf{CCZ} gates in parallel. For asymptotically good qLDPC families, this reduces the existence of a native \mathsf{CCZ} magic-state fountain to a concrete combinatorial problem about the distribution of logical X operators. CSS codes, triple overlaps, and magic-friendly triples. Let C_{X},C_{Z}\subseteq\mathbb{F}_{2}^{n} be binary linear codes with C_{Z}\subseteq C_{X}^{\perp} . A CSS code \mathcal{Q} on n qubits is defined by X -type stabilizers from a parity-check matrix of C_{Z} and Z -type stabilizers from a parity-check matrix of C_{X} . The code encodes k=n-\operatorname{rank}H_{X}-\operatorname{rank}H_{Z} (1) logical qubits, and its X -type logical operators correspond to cosets in \mathcal{L}_{X}=C_{Z}^{\perp}/C_{X}, (2) where C_{Z}^{\perp} denotes the orthogonal complement of C_{Z} with respect to the standard inner product on \mathbb{F}_{2}^{n} . We call \mathcal{Q} a qLDPC code if both H_{X} and H_{Z} have bounded row and column weights, independent of n . A family \{\mathcal{Q}_{n}\} is asymptotically good if k_{n}\geq Rn and d_{n}\geq\delta n for constants R,\delta>0 . For x,y,z\in\mathbb{F}_{2}^{n} , define the triple overlap \tau(x,y,z)=\sum_{i=1}^{n}x_{i}y_{i}z_{i}\pmod{2}, (3) which counts modulo 2 the positions where all three vectors have a 1. Consider three unencoded computational basis strings x,y,z\in\{0,1\}^{n} and |\psi\rangle=\bigotimes_{i=1}^{n}|x_{i}\rangle\otimes|y_{i}\rangle\otimes|z_{i}\rangle. (4) Applying a physical \mathsf{CCZ} gate to each triple (i,i,i) multiplies |\psi\rangle by (-1)^{\tau(x,y,z)} ; the proof is given in Appendix A . Thus the triple overlap completely controls the phase imparted by a wirewise \mathsf{CCZ} layer. When x,y,z represent X -type logical operators (i.e. x,y,z\in C_{Z}^{\perp} ), this relation is the starting point for diagonal logical \mathsf{CCZ} gates 4 ; 12 ; 15 ; 28 . We isolate a simple algebraic condition on triples of logical X operators that is necessary (and in many constructions close to sufficient) for a nontrivial logical \mathsf{CCZ} action. Definition 1 (Magic-friendly triple) . Let \mathcal{Q} be a CSS code with C_{Z}^{\perp}\subseteq\mathbb{F}_{2}^{n} . A triple (x,y,z)\in(C_{Z}^{\perp})^{3} is magic-friendly if: 1. x,y,z represent three independent logical X operators, i.e., their images in \mathcal{L}_{X} are linearly independent; 2. pairwise orthogonality holds: \langle x,y\rangle=\langle x,z\rangle=\langle y,z\rangle=0\pmod{2}; (5) 3. the triple overlap is odd: \tau(x,y,z)=1\pmod{2}. (6) Remark 1 (Logical Z corrections) . In more general CSS constructions, one may allow \langle x^{(a)},x^{(b)}\rangle\neq 0 for some pairs of logical X operators, in which case a transversal application of \mathsf{CCZ} induces additional logical Z -type Clifford corrections on spectator logical qubits. This is the familiar situation in triorthogonal and generalized divisible codes 6 ; 14 . In our setting we impose \langle x,y\rangle=\langle x,z\rangle=\langle y,z\rangle=0 for all participating logicals in a magic-friendly triple. As a result, the only non-trivial logical phase comes from the odd triple overlap, and a depth-one layer of physical \mathsf{CCZ} on representatives of x,y,z acts as the desired logical \mathsf{CCZ} on the three chosen