Author: Rowan Quni-Gudzinas (QNFO/QWAV) | Date: 2026-06-04 | License: QNFO Unified License Agreement (QNFO-ULA)

1. Introduction

Quantum error correction (QEC) is the central organizing principle of quantum computing. The threshold theorem [established] guarantees that if physical error rates fall below a constant threshold, arbitrarily long quantum computations become possible with polynomial overhead. This theorem transformed QEC from theory into an engineering program backed by billions of dollars of investment.

And yet: after three decades, no one has demonstrated a logical qubit that outperforms physical qubits in a scalable, reproducible way [established]. Current physical qubit error rates sit at $10^{-4}$ to $10^{-3}$, while useful large-scale computations may require logical failure rates of $10^{-9}$ or lower — a five-to-six-orders-of-magnitude gap [Spencer et al., arXiv:2605.29137, 2026].

This paper proposes a structural explanation: conventional QEC is built on five hidden Archimedean assumptions, and if the actual metric of quantum error spaces is non-Archimedean (ultrametric), then the mathematical framework used to define, construct, and decode quantum codes is mismatched to the physical reality it aims to control. We call this the metric mismatch hypothesis.

2. The Five Archimedean Assumptions of Conventional QEC

Assumption 1: The Hamming Metric. Code distance $d$ is measured by Hamming distance, which satisfies the Archimedean triangle inequality $d(x,z) \leq d(x,y) + d(y,z)$. This makes “locality” meaningful in error correction.

Assumption 2: Euclidean Locality. Topological codes and the threshold theorem assume errors are local — affecting only immediate neighbors in a spatial layout.

Assumption 3: The Classical-Quantum Cut. The syndrome is treated as a classical bit-string, assuming clean separation between quantum system and classical measurement.

Assumption 4: Markovian, Archimedean-Temporal Noise. Standard noise models assume memoryless environments with constant error rate $p$ per gate.

Assumption 5: Linear Algebra Over Archimedean Fields. Quantum mechanics is formulated over $\mathbb{C}$; stabilizer formalism operates over finite fields $\mathbb{F}_2$, $\mathbb{F}_q$.

3. p-adic Mathematical Preliminaries

Let $p$ be a prime. The $p$-adic absolute value is $|x|_p = p^{-v_p(x)}$, where $v_p(x)$ is the exponent of the highest power of $p$ dividing $x$. This absolute value is non-Archimedean: $|x+y|_p \leq \max(|x|_p, |y|_p)$.

The $p$-adic numbers $\mathbb{Q}p$ admit digit expansions $x = \sum{i=v_p(x)}^{\infty} a_i p^i$ with $a_i \in {0,\ldots,p-1}$. For finite approximations we use $\mathbb{Z}/p^n\mathbb{Z}$.

Ultrametric spaces are isometric to the leaves of a rooted tree, with distance measured by the height of the lowest common ancestor. Every point in an open ball is its center; two open balls are either disjoint or nested.

4. p-adic Stabilizer Codes

4.1 p-adic Pauli Operators

On $\mathcal{H}_p = L^2(\mathbb{Z}_p, d\mu)$ with Haar measure, define shift and clock: $X|x\rangle = |x+1\rangle$, $Z|x\rangle = \chi_p(x)|x\rangle$, where $\chi_p(x) = e^{2\pi i {x}_p}$. These satisfy $XZ = e^{2\pi i/p} ZX$.

For finite register $\mathbb{C}^{p^n}$: $X|x\rangle = |x+1 \bmod p^n\rangle$, $Z|x\rangle = \omega_p^x|x\rangle$, $\omega_p = e^{2\pi i/p^n}$.

4.2 p-adic Weight and Distance

The $p$-adic weight of a single-register Pauli is $|X^a Z^b|_p = \max(|a|_p, |b|_p)$. For multi-register $P = \bigotimes_j X^{a_j} Z^{b_j}$: $|P|_p = \max_j \max(|a_j|_p, |b_j|_p)$.

The $p$-adic code distance is $d_p = \min_{L \in \mathcal{N}(\mathcal{S})\setminus\mathcal{S}} |L|_p$. Smaller $d_p$ means better protection.

4.3 p-adic Concatenation Theorem

Theorem [conjecture]. A $p$-adic stabilizer code over registers of dimension $p^n$ admits a natural concatenated decomposition $\mathcal{C} = \mathcal{C}0 \succ \mathcal{C}_1 \succ \cdots \succ \mathcal{C}{n-1}$ where each $\mathcal{C}_j$ is a code over registers of dimension $p^{n-j}$.

Proof sketch. The projection $\mathbb{Z}/p^n\mathbb{Z} \to \mathbb{Z}/p^{n-1}\mathbb{Z}$ induces a projection on the Pauli group. Iterating gives the chain. $\square$

5. p-adic Quantum Bounds [conjecture]

p-adic Hamming Bound: Sum of $p$-adic error ball volumes cannot exceed total Hilbert space dimension, with error counting organized by $p$-adic weight.

p-adic Singleton Bound: $n - k \geq 2(d_p - 1)_p$ where $(\cdot)_p$ is the $p$-adic ceiling, differing from the standard bound in scaling behavior.

p-adic Gilbert-Varshamov: Existence of $p$-adic codes with rate exceeding a $p$-adic entropy threshold.

6. p-adic Decoders [speculative]

$p$-adic belief propagation generalizes standard BP: messages are $p$-adic probability distributions over $\mathbb{Z}/p^n\mathbb{Z}$. The tree-propagation property of ultrametric spaces suggests convergence in $O(\log_p n)$ iterations, compared to $O(n)$ for standard binary BP. This connects to renormalization-group decoders, which already exploit hierarchical coarse-graining — the $p$-adic framework provides their mathematical foundations.

7. p-adic qLDPC Constructions [conjecture]

$p$-adic hypergraph product: given classical $p$-adic codes with parity-check matrices $H_1, H_2$ over $\mathbb{Z}/p\mathbb{Z}$, the $p$-adic HGP yields a CSS code with matrices over $\mathbb{Z}/p\mathbb{Z}$.

$p$-adic lifted product: replacing cyclic group algebra $\mathbb{F}2[C\ell]$ with $p$-adic group algebra $\mathbb{Z}/p\mathbb{Z}[G]$ for non-abelian $G$ may achieve $k = \Theta(n)$, $d_p = \Theta(n)$ — constant rate with growing $p$-adic distance.

8. The Metric Mismatch Hypothesis

Hypothesis [my conjecture]: The persistent failure to demonstrate scalable QEC arises because conventional codes assume Archimedean error structure, while the actual error structure of quantum systems is ultrametric. The mismatch between code metric and error metric imposes overhead growing with hierarchy depth.

Thermodynamic argument [speculative]: In the Archimedean framework, Landauer cost of erasing syndrome bits is polynomial. In the ultrametric framework, if errors have $L$ hierarchical levels, the total cost scales as $E_{\text{total}} \sim \sum_{\ell=1}^L N_\ell k_B T \ln 2$. For a binary tree, $N_\ell = 2^\ell$, giving $E_{\text{total}} \sim 2^{L+1} k_B T \ln 2$ — exponential in hierarchy depth.

If this scaling holds, error correction becomes thermodynamically infeasible at the scale required for Shor’s algorithm, regardless of encoding rate.

9. The Feynman-Shor Bifurcation

The ultrametric perspective reframes a central tension: Feynman (1982) envisioned quantum simulation; Shor (1994) weaponized quantum computers against cryptography. Simulation tasks naturally have hierarchical (RG-flow) structure matching ultrametric error structure, reducing effective QEC overhead. Cryptanalytic tasks live in Archimedean geometry, mismatched with ultrametric errors.

Testable prediction: The first useful quantum computer will run simulation, not cryptanalysis. Disconfirmed if RSA-2048/secp256k1 are broken before a classically-intractable quantum simulation is demonstrated.

10. Experimental Protocol [speculative]

To test for ultrametric error structure on existing hardware:

  1. Run repeated rounds of surface-code syndrome extraction without active correction. Collect $R$ rounds of $n_s$ syndrome bits.
  2. Compute pairwise error correlations $C(j,k)$ between all checks.
  3. Regress $C(j,k)$ against Euclidean distance $d_E(j,k)$ and $p$-adic distance $d_p(j,k)$.
  4. Compute the ultrametricity index $\mathcal{U}_p$: partial $R^2$ of $d_p$ after controlling for $d_E$. Under Archimedean null: $\mathcal{U}_p \approx 0$. Under ultrametric alternative: $\mathcal{U}_p > 0$ for some $p$.

Feasible with $d \approx 10$–$20$, $R \approx 10^4$–$10^5$, and $O(n_s^2 R)$ classical post-processing.

11. Conclusion

We have proposed that Archimedean assumptions may be the root cause of QEC’s failure to scale. We developed the mathematical framework for $p$-adic stabilizer codes, showed automatic hierarchical concatenation, and formulated the metric mismatch hypothesis with falsifiable predictions.

Three paths forward: (A) Formal theory — prove $p$-adic bounds and decoder convergence. (B) Experimental test — measure $\mathcal{U}_p$ on quantum hardware. (C) Hardware co-design — $p$-adic-native qubit layouts and syndrome extraction.

The hypothesis, if confirmed, reframes QEC from an engineering challenge of perfecting Archimedean codes to a foundational physics challenge of identifying the correct metric of quantum error spaces.

Certainty: Sections 1–4 are [established]. Concatenation theorem [conjecture]. Bounds [conjecture]. Decoders [speculative]. Thermodynamic argument [speculative]. Metric mismatch hypothesis [my conjecture, falsifiable]. Experimental protocol [speculative].