Mode multiplexing for scalable cavity-enhanced operations in neutral-atom arrays

Basic Information

• Paper ID: 2511.20858
• Title: Mode multiplexing for scalable cavity-enhanced operations in neutral-atom arrays
• Authors: Ziv Aqua, Matthew L. Peters, David C. Spierings, Guoqing Wang, Edita Bytyqi, Thomas Propson, Vladan Vuletić (MIT)
• Classification: quant-ph (Quantum Physics)
• Publication Date: November 27, 2025
• Paper Link: https://arxiv.org/abs/2511.20858

Abstract

Neutral-atom arrays provide a versatile platform for quantum information processing. However, in large-scale arrays, efficient photon collection remains a bottleneck for critical tasks such as fast non-destructive qubit readout and remote entanglement distribution. This paper proposes a cavity-based approach that leverages multiple modes of a single optical cavity to enable fast parallel operations on multiple atoms. By selectively shifting relevant atomic transitions, each atom can be coupled to different cavity modes, enabling independent synchronized processing. The practical system design proposed supports cavity-mode multiplexing (CMM) with up to 50 modes, enabling fast mid-circuit syndrome extraction and significantly improving entanglement distribution rates between remote atom arrays. This method provides a scalable solution to core challenges in neutral-atom arrays, advancing practical quantum technology development.

Research Background and Motivation

1. Core Problems

Neutral-atom arrays, as a leading platform for fault-tolerant quantum computing, face bottlenecks in two critical hardware-level operations:

• Non-destructive qubit readout: Used for mid-circuit measurements (MCM) in quantum error correction (QEC) codes, enabling entropy removal
• Remote entanglement distribution: Enabling scalability through interconnected quantum processing module networks

2. Problem Significance

• Practical-scale atomic quantum processors are expected to require millions of qubits operating at high-speed QEC cycles
• Current systems use high-numerical-aperture microscope objectives for photon collection, but collection efficiency is limited (≲10%)
• Non-destructive readout and remote entanglement generation are typically limited to millisecond timescales, becoming bottlenecks for QEC cycle rates

3. Limitations of Existing Approaches

Free-space approaches:

• Low photon collection efficiency (~10%)
• Long readout times (~5ms)
• Low remote entanglement generation rates (~250Hz)
• Stray photon scattering on data qubits causes decoherence

Single-mode cavity approaches:

• Operations must be executed serially, with only one atom interacting with the cavity at a time
• Fundamental trade-off: small cavity mode size increases cooperativity but limits the number of coupled atoms
• Small cavities require atom transport in and out, introducing time overhead
• Rydberg gate performance in compact cavities may be affected by stray electric fields

Multi-cavity array approaches:

• Introduce new challenges such as limited optical access and reduced cavity performance
• High implementation complexity

4. Research Motivation

Need for a scalable cavity integration method that can simultaneously:

• Maintain high cooperativity advantages
• Accommodate large numbers of atoms
• Maintain sufficient distance between mirrors and atoms
• Remain compatible with neutral-atom array geometry and other constraints

Core Contributions

1. Proposes the Cavity-Mode Multiplexing (CMM) concept: Utilizes multiple longitudinal and transverse modes of a single optical cavity, selectively coupling atoms to non-degenerate cavity modes through local light shifts, achieving frequency and spatial multiplexing
2. Designs a system for fast syndrome extraction:
   • Supports 50 cavity modes (25 Hermite-Gauss transverse modes × 2 longitudinal modes)
   • Can simultaneously couple up to 6400 atoms
   • Achieves approximately two orders of magnitude speedup compared to free-space imaging
   • Readout time for 5000 syndrome qubits is only 50μs (vs 5ms for free-space)
3. Designs a system for high-speed remote entanglement:
   • Supports 15 cavity modes (Laguerre-Gauss radial modes)
   • Achieves 4MHz atomic-atomic Bell state generation rate
   • Complete sequence of 40 remote CNOT gates can be performed in ~70μs
   • Compatible with cavity-based Rydberg gates
4. Provides detailed theoretical analysis and system parameters: Including cooperativity calculations, photon generation probabilities, crosstalk suppression, frequency stability, and other key performance metrics

Methodology Details

Task Definition

This paper investigates two core tasks:

Task 1: Fast Syndrome Extraction

• Input: Atom array containing thousands of syndrome qubits
• Output: Identification of syndrome qubits in the |1⟩ state (errors)
• Constraints: Completion on microsecond timescale, low crosstalk, high fidelity

Task 2: High-speed Remote Entanglement Distribution

• Input: Communication atoms in two remote modules
• Output: Atomic-atomic Bell states
• Constraints: High generation rate, high fidelity, support for cavity-based Rydberg gates

Model Architecture

1. Basic Physical Configuration

Atomic energy level structure (using ⁸⁷Rb as example):

• Ground state |g⟩
• Excited state |e⟩ (coupled to cavity modes)
• Highly excited state |f⟩ (used for light shift control)

Optical cavity: Fabry-Pérot cavity supporting:

• Longitudinal modes: spaced by free spectral range (FSR)
• Transverse modes: Hermite-Gauss (HG) or Laguerre-Gauss (LG) modes

Control mechanism:

• Control beam drives |e⟩→|f⟩ transition with Rabi frequency Ωc and detuning Δc
• Resonant control (Δc=0) produces dressed states |e±⟩ = (|e⟩±|f⟩)/√2
• Ground-to-dressed-state transition frequency is ∓Ωc/2 relative to original transition

2. Syndrome Extraction Architecture

Cavity parameters:

Length: 25mm
Mode waist: 20μm
Rayleigh range: 1.6mm
Finesse: 6×10⁴
FSR: 6GHz
Linewidth (FWHM): 100kHz

Mode configuration:

• Uses HGn,0 modes (n∈0,24)
• Transverse modes equally spaced within FSR at 240MHz intervals
• Spanning 2 FSRs, total of 50 modes
• Total frequency range: 12GHz

Atom array layout:

• Atoms arranged along z-axis at intensity maxima of various HG modes
• Odd (even) n-value modes assigned to positive (negative) x-axis
• Minimum inter-atom spacing: ~4μm
• Total capacity: 6400 atoms (50 registers × 128 atoms)

Readout scheme:

• Qubit states: |0⟩≡|5S₁/₂, F=1, mF=1⟩, |1⟩≡|5S₁/₂, F=2, mF=2⟩
• Excited state: |e⟩≡|5P₃/₂, F'=3, mF=3⟩
• Control state: |f⟩≡|4D₅/₂, F''=4, mF=4⟩
• Probe Rabi frequency: Ωp/(2π)=15MHz
• Probe detuning: Δp/(2π)=120MHz
• Average cooperativity: η̄=6.8, standard deviation ση=1.9

Adaptive search protocol: For a batch of nbatch atoms, expected number of queries: $m = 1 + n_{batch}p_{synd}[1 + \log_2(n_{batch})]$

Total readout steps (using nmodes modes): $\text{steps} = \lceil\frac{N}{n_{batch}n_{modes}}\rceil \times m$

3. Remote Entanglement Architecture

Cavity parameters:

Length: 6.25mm
Mode waist: 8.6μm
Rayleigh range: 0.3mm
Finesse: 10⁴
FSR: 24GHz
Linewidth (FWHM): 2.34MHz

Mode configuration:

• Uses LGp,0 radial modes (p∈0,32)
• Modes equally spaced within FSR at 727MHz intervals
• Peak intensity at (x,y)=(0,0) independent of p
• Available modes: 15 (excluding modes affected by residual light shifts)

Atom array layout:

• One-dimensional array along z-axis, 225 atoms
• Inter-atom spacing: ~3μm
• Divided into 15 registers, each with 15 atoms
• Average cooperativity: η̄=7, standard deviation ση=1.8

vSTIRAP protocol:

• Qubit states: |0⟩≡|5S₁/₂, F=1, mF=-1⟩, |1⟩≡|5S₁/₂, F=1, mF=1⟩
• Initial state: |g⟩≡|5S₁/₂, F=2, mF=0⟩
• Excited state: |e⟩≡|5P₃/₂, F'=1, mF=0⟩
• Control state: |f⟩≡|4D₅/₂, F''=2, mF=0⟩

Photon generation probability: $\alpha_{interface} = \frac{\eta}{\eta+1} \times \frac{\kappa_e}{\kappa_e+\kappa_i}$

Atomic-atomic entanglement success probability: $P_s = \frac{1}{2}(\alpha_{interface} \times \alpha_{setup})^2$

where κe/(2π)=2.3MHz (output coupling rate), κi/(2π)=38kHz (internal loss rate), αsetup=0.75 (optical system efficiency)

Technical Innovations

1. Local light shift control:
   • Utilizes excited-state-to-excited-state transition-induced light shifts
   • Maximum frequency shift up to 2000Γ (approximately 12GHz for ⁸⁷Rb)
   • Frequency fluctuations limited to natural linewidth under 0.1% intensity stability
2. Mode separation techniques:
   • Transverse modes: Multi-plane light conversion (MPLC), scalable to >1000 modes
   • Longitudinal modes: Virtual imaging phase array (VIPA), frequency resolution Δf/f∼10⁶
3. Adaptive batch processing:
   • Leverages statistical properties of low error probability
   • Global check + binary search
   • Logarithmic complexity O(log₂nbatch)
4. Crosstalk suppression:
   • Individual probe addressing
   • Cavity mode frequency separation
   • Crosstalk <10⁻⁵
5. Temperature robustness:
   • Accounts for finite spatial wavefunction at 10μK atomic temperature
   • Reduced cooperativity of higher-order LG modes incorporated into design

Experimental Setup

Dataset

This paper is a theoretical proposal and does not involve experimental datasets. All analyses are based on:

• Known physical parameters of ⁸⁷Rb atoms
• Standard characteristics of optical cavities
• Numerical simulations (e.g., vSTIRAP processes)

Evaluation Metrics

Syndrome extraction:

• Readout time (μs)
• Crosstalk probability
• Photon collection efficiency (>80%)
• Cooperativity η
• Number of queries

Remote entanglement:

• Bell state generation rate (Hz)
• Success probability Ps
• Fidelity
• Complete operation sequence time
• Photon generation probability

Comparison Methods

1. Free-space imaging:
   • Readout time: 5ms
   • Photon collection efficiency: ~10%
2. Single-mode cavity:
   • Serial operation
   • Requires atom transport
3. Micro-cavity arrays:
   • Single-cavity rate: ~2.4MHz
   • Requires atom transport for readout and Rydberg gates

Implementation Details

Syndrome extraction:

• Batch size: nbatch=256
• Query time: 10μs
• Error probability: psynd=5×10⁻³
• Scattered photons per atom: ~20 (within 10μs)

Remote entanglement:

• Control beam switching time: 100ns
• Scanning time: 12μs (15 registers)
• Rydberg gate + transport: ~20μs
• Measurement time: 25μs (40 communication qubits)

Experimental Results

Main Results

1. Syndrome Extraction Performance

Speed improvement:

• 5000 syndromes: CMM 50μs vs free-space 5ms
• ~100-fold improvement

Scalability:

• Single-mode cavity enters millisecond regime at ~2000 atoms
• CMM maintains microsecond operations at 6400 atoms

Photon collection:

• Efficiency: >80%
• Scattered photons per atom: ~20 (10μs)
• Crosstalk: <10⁻⁵

Cooperativity distribution:

• Average η̄=6.8
• Standard deviation ση=1.9
• Normalized to peak cooperativity of HG₀,₀ mode η₀

2. Remote Entanglement Performance

Generation rate:

• Atomic-atomic Bell states: 4MHz
• Far exceeds free-space ~250Hz

Success probability:

• Theoretical limit: Ps=0.21
• Practical average: P̄s=0.18 (accounting for non-adiabatic effects and η distribution)

Complete operation time (40 remote CNOT gates):

• Initialization: 16μs
• Bell pair generation: 12μs
• Rydberg gates: 20μs
• Measurement: 25μs
• Total: ~70μs
• ~100-fold improvement over free-space

Robustness:

• 6MHz atomic detuning causes <1% probability change
• Spontaneous decay to wrong state: <1%
• Applicable in magnetic field environments (Zeeman shifts)

Ablation Studies

While the paper does not conduct traditional ablation experiments, it demonstrates the importance of each component through parameter analysis:

1. Effect of mode number:
   • Figure 3 shows readout time scaling with mode number
   • 50 modes provide significant advantage over single mode
2. Effect of cooperativity:
   • Figure 2(d) shows cooperativity distribution across different HG modes
   • Figure 4(d) shows temperature effect on LG mode cooperativity
3. Effect of vSTIRAP driving:
   • Figure 5 shows effect of different probe field slopes on photon generation
   • Trade-off between photon generation probability and rate

Case Studies

Case 1: Distance-20 surface code QEC cycle

• Requirement: ~40 remote CNOT gates/cycle
• CMM implementation:
  • Communication qubits remain at fixed positions
  • Single scan generates required Bell pairs
  • Cavity-based Rydberg gates reduce atom transport
  • Total cycle time: 70μs

Case 2: Syndrome extraction for 6400-atom array

• Configuration: 50 batches × 128 atoms
• Error rate: psynd=5×10⁻³
• Expected errors: ~32
• Adaptive search effectively locates errors
• Total time: <100μs

Experimental Findings

1. Logarithmic scaling advantage: Logarithmic complexity of adaptive search provides clear advantage in large-scale systems
2. Mode-dependent cooperativity:
   • HG modes: Higher-order modes show reduced cooperativity
   • LG radial modes: Peak cooperativity independent of p
3. Temperature sensitivity: Higher-order LG modes more temperature-sensitive, but still acceptable at 10μK
4. Frequency stability: Stability of dressed-state frequency critical for practical implementation
5. Crosstalk management: Effectively suppressed through mode separation and frequency spacing

Related Work

1. Neutral-Atom Quantum Computing

High-fidelity gate operations:

• Rydberg gate fidelity >99.5% 2-6
• Coherent transport enables arbitrary connectivity 7,8
• Fault-tolerant logical operations demonstrated 9,10

Large-scale systems:

• 6100 highly coherent atoms 11
• 3000 atoms in continuous operation 12-14

2. Cavity-Enhanced Operations

Non-destructive readout:

• Fluorescence detection 31,33,34
• Transmission detection 32,33,35,78
• Microsecond readout times (η>1)

Atom-photon entanglement:

• Neutral atom systems 36-44
• Trapped ion systems 45-51
• vSTIRAP schemes 36,98

Remote entanglement:

• Basic quantum networks 41
• Long-distance entanglement (>50km) 46,92
• Quantum logic gates 44

3. Multi-Mode Cavity Applications

Existing research:

• Single-atom motion tracking 58,59
• Tunable range atom-atom interactions 60,61
• Laughlin photon states 62
• Cooperativity enhancement 63

Distinction of this work:

• First systematic proposal for selective coupling to non-degenerate modes
• Combines frequency and spatial multiplexing
• Large-scale parallel operations

4. Modular Architecture

Theoretical proposals:

• Distributed computing based on small quantum registers 21
• Freely scalable techniques 23
• Atom storage and photonic interconnects 24

Experimental implementations:

• Distributed computing over optical network links 27
• Operations between remote quantum processors

Advantages of this work:

• Higher entanglement generation rates
• Compatible with Rydberg gates
• Reduced atom transport requirements

Conclusions and Discussion

Main Conclusions

1. CMM provides scalable solution: By leveraging multiple modes of a single cavity, CMM overcomes the serial operation bottleneck of single-mode cavities while avoiding complexity of multi-cavity systems
2. Significant performance improvements:
   • Syndrome extraction: ~100-fold improvement over free-space
   • Remote entanglement: 4MHz generation rate, far exceeding existing methods
3. Practical system design: Paper provides detailed parameters and configurations supporting 50-mode multiplexing, accommodating thousands of atoms
4. Broad applicability: CMM concept not limited to ⁸⁷Rb, applicable to other alkali, alkaline-earth atoms, and ions

Limitations

1. Stringent hardware requirements:
   • Requires precise cavity symmetry control (cylindrical or intentional breaking)
   • Requires 3D piezoelectric translation stages for mirror control
   • Individual atom addressing capability still under development
2. Light shift control challenges:
   • Requires 0.1% intensity stability
   • Beam pointing fluctuations and finite atomic temperature may cause heating
   • Two-dimensional individual control requires advanced techniques 79-85
3. Temperature sensitivity: Higher-order LG modes sensitive to atomic temperature, requiring effective cooling
4. Crosstalk management: Tight spacing requires register grouping and sequential scanning
5. Theoretical proposal: Lacks experimental verification; actual performance may be affected by unforeseen factors

Future Directions

1. Alternative cavity geometries:
   • Bowtie cavities: Small mode waist + long cavity length, high cooperativity + small FSR
   • Reduce dependence on high-order spatial modes
2. Optimized driving protocols: vSTIRAP parameters and driving profiles can be further optimized to improve generation rates
3. Extended applications:
   • Other protocols in quantum networks
   • Quantum sensing
   • Quantum simulation
4. Integrated photonics:
   • Integrated photonic chips for atom addressing 81,83
   • On-chip mode separation
5. Adaptive protocols:
   • Priority syndrome extraction 100
   • Dynamic resource allocation

In-Depth Evaluation

Strengths

1. Strong innovation:
   • First systematic proposal of cavity-mode multiplexing for large-scale atom arrays
   • Clever combination of frequency and spatial multiplexing
   • Selective coupling via local light shifts
2. Rigorous theory:
   • Detailed physical models and numerical calculations
   • Accounts for dressed states, temperature effects, crosstalk, and other practical factors
   • Provides complete system parameters
3. High practical value:
   • Addresses critical bottlenecks in fault-tolerant quantum computing
   • ~100-fold improvement over existing methods
   • Compatible with existing technologies
4. Comprehensive design:
   • Detailed designs for two different application scenarios
   • Considers geometric constraints, Rydberg gate compatibility, and other practical issues
   • Provides specific technical solutions for mode separation
5. Clear presentation:
   • Logical structure
   • Rich and informative figures
   • Sufficient technical detail

Weaknesses

1. Lack of experimental verification:
   • Pure theoretical proposal without experimental data
   • Actual implementation may face unforeseen technical challenges
2. High hardware requirements:
   • Individual atom addressing technology still under development
   • Precision cavity control technology demanding
   • May limit near-term implementation
3. Insufficient parameter optimization:
   • vSTIRAP driving parameters could be further optimized
   • Parameter space not fully explored
4. Missing cost analysis:
   • No discussion of implementation costs
   • Insufficient cost comparison with multi-cavity approaches
5. Limited error analysis:
   • Lacks detailed analysis of cumulative error effects
   • Fidelity estimates may be overly optimistic

Impact

1. Contribution to field:
   • Provides new pathway for large-scale neutral-atom quantum computing
   • Likely to influence future quantum processor architecture design
   • Promotes deep integration of cavity QED with atom arrays
2. Practical value:
   • Direct application to QEC cycle acceleration
   • Supports modular quantum computing architectures
   • Potential applications in quantum networks
3. Reproducibility:
   • Provides detailed system parameters
   • Clear physical principles
   • Requires advanced experimental techniques for actual implementation
4. Inspirational value:
   • CMM concept generalizable to other systems (ions, molecules, etc.)
   • Mode multiplexing ideas applicable to other quantum information tasks
   • May inspire new hybrid quantum system designs

Applicable Scenarios

1. Quantum error correction:
   • QEC schemes requiring fast syndrome extraction (surface codes, etc.)
   • Large-scale quantum processors (>1000 qubits)
2. Modular quantum computing:
   • Interconnection of multiple atom array modules
   • Distributed quantum computing architectures
3. Quantum networks:
   • High-speed entanglement distribution between nodes
   • Quantum repeaters
4. Hybrid systems:
   • Atom-photon interfaces
   • Quantum sensing networks
5. Less suitable scenarios:
   • Small-scale systems (<100 qubits): complexity not justified
   • Applications not requiring high-speed operations
   • Scenarios with severely limited hardware resources

References

This paper cites extensive important literature, with key references including:

1. Neutral-atom quantum computing 1-14: High-fidelity gates, coherent transport, fault-tolerant operations
2. Cavity QED fundamentals 30-35: Purcell effect, non-destructive readout
3. Atom-photon entanglement 36-51: vSTIRAP, remote entanglement
4. Mode separation techniques 70-77: MPLC, VIPA
5. Modular architectures 20-27: Distributed quantum computing theory and experiments

Overall Assessment: This is a high-quality theoretical proposal paper that introduces an innovative cavity-mode multiplexing concept, providing a promising solution for large-scale neutral-atom quantum computing. The paper is theoretically rigorous, comprehensively designed, and thoroughly analyzed, with potential for significant impact on quantum information science. The main challenge lies in experimental implementation complexity, but with advances in individual atom addressing and related technologies, this approach shows strong feasibility.