⚛️ quantum physics

Coprime Bivariate Bicycle Codes and Their Layouts on Cold Atoms

This paper introduces a novel subclass of bivariate bicycle quantum error correction codes based on coprime polynomials that enable predictable code rates and the discovery of new short-to-medium length codes, alongside a specialized cold atom layout that significantly reduces movement time and operations for syndrome extraction under global laser noise.

Original authors: Ming Wang, Frank Mueller

Published 2026-03-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ming Wang, Frank Mueller

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to send a secret message across a stormy ocean. The waves (noise) are huge, and your boat (the quantum computer) is fragile. If you just send one small message, a single wave could sink it. To survive, you don't just send one boat; you build a massive, redundant fleet where the message is split up and hidden in many different places. If some boats get hit by waves, the others can still reconstruct the original message. This is the essence of Quantum Error Correction.

However, building this fleet is expensive. You need a lot of extra boats (physical qubits) just to protect one piece of information (a logical qubit). The goal of this paper is to design a better, more efficient fleet that uses fewer resources and is easier to navigate, specifically for a type of ship called a Cold Atom Array.

Here is a breakdown of the paper's key ideas using everyday analogies:

1. The Problem: The "Vanilla" Recipe is Too Random

Previously, scientists designed these error-correcting codes (the fleet plans) using a "Vanilla" recipe called Bivariate Bicycle (BB) codes.

The Analogy: Imagine trying to build a perfect puzzle by randomly grabbing pieces and hoping they fit. You might find a great puzzle, but you don't know how big it will be or how many pieces you need until you've already started building.
The Issue: This trial-and-error approach is slow. You can't plan your budget (the number of qubits) in advance because the "rate" (how much actual data you can store) is a mystery until the code is finished.

2. The Solution: The "Coprime" Blueprint

The authors, Ming Wang and Frank Mueller, invented a new way to build these codes called Coprime-BB codes.

The Analogy: Instead of randomly grabbing puzzle pieces, they decided to use a specific rule: "We will only use pieces that fit together in a pattern where the number of rows and columns share no common factors (coprime)."
The Benefit: By following this strict rule, they can predict the size of the puzzle before they even start building it. They can say, "We want a fleet that holds 12 messages; we know exactly how many extra boats we need to build it." This makes the design process much faster and more reliable.

3. The Hardware: The "Dancing Atoms"

The paper focuses on Cold Atom Arrays.

The Analogy: Imagine a dance floor where every dancer is an atom. To check if a dancer made a mistake (an error), they need to pair up and high-five (interact). In a standard layout, dancers are arranged in a square grid. To check everyone, the dancers have to shuffle around in a complex, 2D pattern, crossing over each other many times.
The Old Way (BB Layout): To check a specific pair, dancers might have to move left, then up, then right, then down. Every time they move, there's a risk of tripping (noise) or getting hit by a global spotlight (laser noise) that affects everyone on the floor.

4. The Innovation: The "CBB Layout" (The Single-File Line)

The authors realized that because their new "Coprime" codes have a special mathematical structure, the dancers don't need to shuffle in a 2D square.

The Analogy: They reorganized the dance floor into a long, single-file line (a 1D strip).
How it works: Instead of moving in a complex 2D grid, the dancers just slide forward or backward in a circle.
The Result:
1. Fewer Moves: They don't have to zigzag. They just slide.
2. Less Noise: Because they move fewer times, they are exposed to the "global spotlight" (laser noise) fewer times.
3. Speed: Even though the line is long, sliding in a straight line is faster than doing a complex 2D shuffle.

5. The Proof: Simulations Show It Works

The authors ran computer simulations (like a flight simulator for their quantum fleet).

The Findings:
- Their new codes (Coprime-BB) found many new, efficient designs that nobody knew existed before.
- When they used their new "Single-File Line" layout (CBB) instead of the old "Square Grid" layout (BB), the error rates dropped significantly.
- The Magic Number: For some codes, the new layout reduced errors by a factor of 10 when the environment was very noisy.

Summary

Think of this paper as an engineer who:

Invented a new blueprint (Coprime-BB codes) that lets you know exactly how much material you need before you start building.
Redesigned the construction site (CBB Layout) so the workers (atoms) only have to walk in a straight line instead of a confusing maze.
Proved that this new method is faster, cheaper, and keeps the final product (the quantum computer) much safer from storms (errors).

This work is a significant step toward making quantum computers practical, moving us from the "noisy, experimental" era to a future where we can build reliable, large-scale quantum machines.

1. Problem Statement

Quantum Error Correction (QEC) is essential for scaling quantum computers, yet current approaches face trade-offs between code performance and hardware feasibility.

Surface Codes: While popular and hardware-friendly (2D grid), they suffer from high overhead (requiring $L^2$ physical qubits for one logical qubit).
qLDPC Codes: These offer better asymptotic parameters (linear scaling of rate and distance) but are difficult to implement on hardware due to complex connectivity requirements.
Bivariate Bicycle (BB) Codes: A promising subclass of qLDPC codes that can be embedded on 2D planes (suitable for cold atom arrays). However, constructing them is computationally expensive because:
1. Parameters (rate $k$ and distance $d$ ) are unknown until after a numerical search is completed.
2. Existing layouts on cold atom arrays require multiple movement steps for "mixed" polynomial terms, increasing exposure to global laser noise and idle qubit errors.

2. Methodology

The authors propose a three-pronged approach: a new code construction method, an accelerated search algorithm, and a hardware-tailored layout.

A. Coprime-BB Code Construction

The authors introduce a novel subclass of BB codes called Coprime-BB codes.

Mathematical Formulation: Instead of using two independent variables $x$ and $y$ separately (as in vanilla BB codes), they define the code using a single variable $\pi = xy$ .
Constraints: The dimensions $l$ and $m$ of the underlying cyclic groups are required to be coprime integers.
Polynomials: The parity-check matrices are defined by univariate polynomials $a(\pi)$ and $b(\pi)$ over the ring $\mathbb{F}_2[\pi]/(\pi^{lm} + 1)$ .
Key Advantage: The code rate $k$ can be determined a priori using the formula $k = 2 \cdot \deg(\gcd(a(\pi), b(\pi), \pi^{lm}+1))$ . This allows researchers to target specific code rates before running the search.

B. Accelerated Search Algorithms

Two algorithms are presented to find optimal codes:

Algorithm 1 (Vanilla BB Search): Optimizes the search for standard BB codes by eliminating equivalent codes (codes with identical $n, k, d$ parameters generated by swapping $A$ and $B$ matrices) and using thresholds to skip low-rate or low-distance candidates early.
Algorithm 2 (Coprime-BB Search): Utilizes the coprime construction. It iterates through factors of $\pi^{lm}+1$ to select a target GCD polynomial (determining $k$ ), then searches for polynomials $a(\pi)$ and $b(\pi)$ that satisfy this GCD. This drastically reduces the search space compared to brute-force methods.

C. CBB Layout Optimization

The authors propose a new physical layout for cold atom arrays, termed the CBB layout, specifically optimized for the mixed-term nature of Coprime-BB codes.

Standard BB Layout: Requires 4 movement steps to execute checks involving mixed terms (e.g., $xy$) because it must cross both vertical and horizontal periodic boundaries.
CBB Layout: Arranges qubits in a 1D strip (or long stripe) ordered by powers of $\pi$ ( $1, \pi, \pi^2, \dots$ ).
Mechanism: Since every monomial in Coprime-BB codes is of the form $\pi^i = (xy)^i$ , syndrome extraction becomes a simple cyclic shift along the 1D strip. This reduces the number of movement steps from 4 to 2.
Noise Model: The simulation incorporates a global Rydberg laser noise model, where every movement step exposes all atoms (including idle ones) to noise, making the reduction in movement steps critical for performance.

3. Key Contributions

Novel Code Construction: Introduction of Coprime-BB codes, which generalize BB codes by allowing mixed terms while enforcing coprime dimensions. This enables pre-determination of the code rate, a significant improvement over previous search methods.
Discovery of New Codes: The algorithms successfully identified several previously unknown short-to-medium length codes with favorable parameters (e.g., $[[126, 12, 10]]$ , $[[154, 6, 16]]$ ).
Hardware-Aware Layout: Development of the CBB layout tailored for cold atom arrays. It exploits the algebraic structure of coprime codes to minimize atom movement.
Comprehensive Simulation: Evaluation of the codes under both code-capacity and circuit-level noise models, specifically accounting for global laser noise on cold atoms.

4. Results

Code Performance:
- The new Coprime-BB codes demonstrate competitive or superior performance compared to existing BB codes and surface codes, particularly in the short-to-medium length regime ( $n < 200$ ).
- The metric $kd^2/n$ (normalized efficiency) shows that these codes offer better trade-offs than surface codes ( $n \propto kd^2$ ).
Layout Efficiency:
- Movement Time: Despite potentially longer individual move distances, the CBB layout reduces the total time per syndrome extraction cycle because it requires fewer acceleration/deceleration phases (fewer "stops").
- Error Rates: Under the circuit-level noise model with global laser noise ( $c=0.1$ to $0.5$), the CBB layout consistently achieves lower logical error rates than the standard BB layout.
- Impact of Noise: For high global noise coefficients ( $c=0.5$ ), the CBB layout reduced logical error rates by approximately 10x for high-distance codes (e.g., $[[126, 12, 10]]$ ) compared to the standard layout. This is attributed to the reduction in global laser pulses required per check.

5. Significance

Bridging Theory and Hardware: This work addresses the "hardware gap" in qLDPC codes. While many qLDPC codes exist theoretically, few are practical for specific hardware. The CBB layout makes Coprime-BB codes highly viable for cold atom quantum computers, which possess all-to-all connectivity but suffer from global noise during atom transport.
Efficiency in Search: By allowing the code rate to be fixed before searching, the methodology accelerates the discovery of codes tailored to specific hardware constraints (e.g., a specific number of available qubits).
Noise Mitigation: The study highlights that minimizing the number of global operations (laser pulses) is as critical as the code's algebraic distance. The CBB layout demonstrates that architectural optimization can yield significant gains in logical fidelity, potentially making these codes candidates for quantum memory in near-term devices.
Future Directions: The paper suggests that while these codes are promising for cold atoms, further work is needed to adapt them to superconducting architectures and to develop fault-tolerant logical gate constructions for this specific code family.