The quantum-advantage resource in multimode OPA light: Identification, optimization, extraction

Abstract: We introduce the notion and reveal remarkable properties of quantum complexity resource contained in a mixed multimode Gaussian state and providing universal quantitative characterization of its quantum advantage. The notion is based on convex optimization, multimode photon number statistics, Hafnian Master Theorem, and #P-hard complexity. We consider pulsed OPAs targeting maximal quantum complexity resource and thousands of multipartite-entangled squeezed modes of output light via nonlinear, spatio-temporally nonadiabatic generation inside OPA and optimized extraction out of OPA. We show that such figure of merit is more realistic than Bloch--Messiah supermodes and guides to multimode OPAs opening new paths to important applications in quantum information science such as generation of 3D cluster states for one-way photonic quantum computing and demonstration of quantum advantage.

• The paper rigorously defines the quantum‐advantage resource and demonstrates its unique connection to #P‐hardness in multimode OPA systems.
• The paper introduces a convex optimization framework to isolate the pure squeezed-core resource, outperforming traditional BM-mode analysis in mixed and lossy conditions.
• The paper validates optimal passive and active extraction strategies, providing practical guidelines for scalable continuous-variable quantum information processing.

Quantum-Advantage Resource in Multimode OPA Light: Formalization, Operational Characterization, and Extraction

Introduction and Context

This work rigorously defines, analyzes, and operationalizes the concept of the "quantum-advantage resource" in multimode optical parametric amplifier (OPA) light, with direct implications for continuous-variable (CV) quantum information processing, quantum computation, and boson sampling (2606.18605). The analysis reveals the genuine quantum resource responsible for the exponential computational complexity distinguishing quantum from classical simulability in mixed, multipartite-entangled, multimode bosonic systems—specifically, those produced by nonadiabatic, nonlinear pulsed OPAs.

The established quantum-advantage resource is quantitated through convex optimization on the mode covariance matrix, with strong emphasis on its direct correspondence to #P-hardness via the Hafnian Master Theorem. This resource fundamentally diverges from standard constructs such as Bloch-Messiah (BM) eigen-squeezed supermodes, which are typically misapplied in the literature, especially in noisy or lossy settings.

Definition, Mathematical Structure, and Resource Dimension

The paper introduces the quantum-advantage resource of a multimode Gaussian state as the minimal pure squeezed-core that, after deconvolving all extractable classical noise (semidefinite classical residue), remains nonclassically non-simulatable by efficient classical algorithms. This is formulated by decomposing the measured covariance matrix $V$ into a "quantum" part $V_q$ and "classical" noise $V_c$ via the convex program:

$$\text{min Tr}\{V_q\} \quad \text{s.t.} \quad V - V_q \geq 0, \; V_q \geq \frac{1}{2}I$$

Crucially, the pure-core property is established: the optimal $V_q$ always represents a pure Gaussian state, so that all its symplectic eigenvalues equal $1/2$ (i.e., it saturates the quantum uncertainty limit), corresponding to the necessary algebraic structure for maximal quantum advantage.

A scalar measure of resource is then defined as

$N_q(V) = 2\;\mathrm{Tr}\{V_q - \frac{1}{2}I\}$

which quantifies the mean photon number in the nonclassically irreducible modes. The operational resource dimension relevant for #P-hard bosonic sampling is

$N_{\mathrm{OA}} = \sum_j \min\{1, \sinh^2 r_j\}$

where $r_j$ are the single-mode squeezings of $V_q$. Only modes with mean occupancy $V_q$0 contribute exponentially in the computational complexity.

Connection to Computational Complexity

The resource, as defined, is the unique figure of merit for photonic quantum advantage. This is formally justified by:

1. The #P-hardness of the matrix hafnian governing the output photon-number statistics; efficient classical simulation becomes impossible when the quantum resource, as measured above, is of sufficient dimension.
2. The direct connection to Toda's theorem, establishing that #P-hardness characterizes the complete quantum complexity of Gaussian CV systems.

The generalization to non-Gaussian states is addressed: higher moments/cumulants do not create quantum advantage beyond the #P-complete Hafnian class unless Gaussian-level complexity is insufficient.

Limitations of the Bloch-Messiah Mode Analysis

A central analytical and numerical claim is that BM supermodes, popular in experimental characterization, systematically overcount the resource in mixed states, especially under significant loss. Only on pure states do the BM and resource-mode bases coincide ($V_q$1). For generic mixed states, the BM decomposition integrates classical noise into its squeezing parameters, inflating apparent resource photon numbers by substantial factors—frequently an order of magnitude or more—relative to the true recoverable quantum resource.

Resource Extraction Algorithms and Extraction Ceilings

For practical quantum information applications, maximally extracting quantum resource from an OPA output is key. The paper rigorously analyzes passive and active extraction strategies:

• Passive extraction: Achieved by an optimal orthogonal-symplectic transformation (unitary interferometer) followed by truncation. The per-budget (number of output channels) extraction is sharply bounded. Extraction interferometers must target the $V_q$2 basis, not the BM basis. The resource extraction algorithm based on $V_q$3 always achieves the maximal possible recoverable resource for any mode budget (up to additivity limits).
• Active extraction: By further squeezing (i.e., feeding the resource-bearing modes into a second OPA), the entire pure-core resource can be mapped into the desired output channels, exceeding any passive extraction ceiling.

Algorithms for (i) Wigner-mode selection, (ii) BM supermode-based extraction, (iii) $V_q$4-aligned (true resource) extraction, and (iv) active hill-climbing and hybrid protocols are mathematically characterized and compared. Numerical experiments robustly demonstrate the superiority of the $V_q$5-aligned extraction, especially as resource depleting mechanisms (loss, pruning, binning) become severe.

Resource Depletion Mechanisms

Quantitative and operational analysis of resource depletion mechanisms reveals:

• Loss in individual channels dramatically suppresses the resource even at moderate transmission reduction ($V_q$6). The Wigner lower bound becomes tight under homogeneous loss, and resource vanishes rapidly with mean transmission.
• Mode tracing (pruning), representing missed or undercollected modes in experiment, can annihilate quantum advantage even when a significant fraction of modes is retained, due to the complex entanglement structure.
• Coarse-grained binning (combining several physical modes per detection channel) reduces resource unless the modes combined have identically matched correlations.

Optimal strategies for mode selection, pruning, and collection are elucidated.

Practical and Experimental Considerations

The formalism directly informs OPA and measurement system design. For maximizing quantum resource:

• The nonlinear OPA process should couple as many modes as possible with strong nonadiabaticity, creating multipartite entanglement over thousands (potentially $V_q$7 or more) modes.
• Extraction optics must be engineered to select resource modes in the $V_q$8-optimal basis.
• Excessive single-mode squeezing is not necessary; moderate squeezing ($V_q$9, $V_c$0–10 dB) suffices when resource dimension and multipartite entanglement are high.
• Proof-of-principle experiments need not sample photon numbers directly; measuring the output covariance and verifying resource dimension $V_c$1 suffices to demonstrate computational quantum advantage over known classical simulation algorithms.

Implications for Quantum Information Science

This work addresses a significant gap in standard experimental and theoretical methodology for multimode quantum photonics—specifically, the failure of BM-mode-based metrics to track the recoverable quantum resource under mixedness and loss. By charting a clear path to verified quantum advantage through resource quantification, state engineering, and optimal extraction, the approach promises immediate application to scalable CV quantum computation, boson sampling, and other quantum-enhanced technologies.

A direct implication is that maximizing overall mode count and controlling multipartite entanglement, not merely maximizing squeezing, will have the highest theoretical and practical impact in photonic quantum information processing.

Conclusion

The paper offers a mathematically rigorous, operational framework for identifying, quantifying, optimizing, and extracting quantum-advantage resource in pulsed multimode OPA systems. The findings invalidate standard BM-mode-based resource metrics in lossy/mixed systems and establish that only the pure-core resource, as defined via the presented convex decomposition, acts as the universal quantifier for computational quantum advantage.

This foundational result is expected to guide future experimental designs, inform development of new algorithms for quantum information tasks in high-dimensional photonic systems, and provide benchmarks for scalable, error-corrected quantum computation with continuous variables.

Reference: "The quantum-advantage resource in multimode OPA light: Identification, optimization, extraction" (2606.18605)