Varun Upreti and Ulysse Chabaud of DIENS, École Normale Supérieure, have identified a new property in continuous-variable quantum circuits, termed that mirrors the role of coherence in discrete-variable systems. Their work establishes symplectic coherence as a necessary component for achieving universal computation in continuous-variable quantum computers, alongside non-Gaussianity and entanglement.

The researchers developed a classical simulation algorithm that efficiently tackles computations with low symplectic coherence, revealing its importance as a limit to classical simulation. This discovery demonstrates that the interplay of symplectic coherence, non-Gaussianity and entanglement in continuous-variable quantum computing parallels the discrete-variable relationship between coherence, magic and entanglement via the Gottesman-Kitaev-Preskill encoding.

Non-Gaussianity and Entanglement Enable Universal CV Computation

The absence of symplectic coherence within continuous-variable (CV) permits efficient classical simulation, establishing this property as a fundamental requirement for achieving computational universality in this approach to quantum computing. Researchers demonstrated that circuits lacking symplectic coherence can be efficiently modeled using conventional computers, highlighting its importance as a resource for quantum advantage. This ability to classically simulate circuits with low symplectic coherence is particularly noteworthy because it defines a clear boundary for what can be achieved with existing classical algorithms.

This correspondence suggests a deeper, underlying principle governing quantum computation regardless of the physical implementation. Specifically, CV entanglement naturally generalizes DV entanglement, while CV non-Gaussianity corresponds to DV magic, a concept recently gaining traction in the literature. “Through the GKP encoding, we showed that the CV resource interplay generalizes the DV resource interplay, with CV entanglement, non-Gaussianity, and symplectic coherence corresponding to DV entanglement, magic, and coherence, respectively,” the paper states.

The researchers’ proof relies on a novel CV path simulation technique based on quadrature back-propagation, allowing them to efficiently trace the evolution of quantum states. This technique helps to understand how the interplay of non-Gaussianity, entanglement, and symplectic coherence drives the complexity of CV computations. The team’s analysis, based on a gate set including cubic phase, orthogonal, and Fourier gates, reveals that each of these resources plays a distinct role in enabling universal quantum computation.

The cubic phase gate introduces non-Gaussianity, the two-mode SUM gate generates entanglement, and the Fourier gate induces symplectic coherence. The symplectic matrix of the SUM gate, in the quadrature basis, confirms its role in inducing entanglement within CV computations. Similarly, the symplectic matrix of the Fourier gate demonstrates its function in generating symplectic coherence. This detailed mapping between CV and DV resources provides a valuable framework for understanding the fundamental requirements for quantum computation across different platforms.

The findings have implications for the design and optimization of future CV quantum computers. Via the Gottesman-Kitaev-Preskill encoding, the researchers show that the interplay of these three continuous-variable quantum resources mirrors the discrete-variable relationship between coherence, magic and entanglement.

GKP Encoding Bridges Discrete and Continuous Variable Systems

The mapping of discrete quantum operations to their continuous-variable counterparts reveals a specific symmetry: the Fourier transform gate, essential for inducing symplectic coherence in continuous-variable systems, corresponds directly to the Hadamard gate responsible for coherence in discrete-variable systems. This correspondence, detailed in recent work, extends beyond single gates, establishing a broader equivalence in the resources needed for universal quantum computation across both paradigms. This parallel is made possible through the Gottesman-Kitaev-Preskill (GKP) encoding, a technique that translates discrete quantum states and operations into the infinite-dimensional Hilbert space of continuous variables.

The GKP encoding allows for the application of established discrete-variable simulation techniques to GKP-encoded continuous-variable computations, but the current work extends beyond this specific case to explore the resources required for general continuous-variable quantum computations. The absence of any one of the three resources, non-Gaussianity, entanglement, or symplectic coherence, results in efficient classical simulation, while their combined presence is sufficient for universal quantum computation, a finding that defines a boundary for classical computational power.

The team’s analysis reveals that the cubic phase gate, the continuous-variable analogue of the discrete-variable T gate, introduces non-Gaussianity, a critical resource for computations beyond those achievable with Gaussian states. Similarly, the two-mode SUM gate, corresponding to the discrete-variable CNOT gate, generates entanglement within the continuous-variable system. This detailed mapping of discrete and continuous resources clarifies the fundamental requirements for quantum advantage in both approaches.

This algorithm’s efficiency hinges on the limited complexity of computations lacking substantial symplectic coherence, effectively defining a threshold beyond which classical simulation becomes intractable. “The interplay of CV non-Gaussianity, entanglement, and symplectic coherence generalizes the resource interplay of DV magic, entanglement, and coherence in GKP-encoded quantum computations,” the authors write, highlighting the overarching principle governing both computational models. These platforms offer enhanced noise robustness and enable high levels of entanglement. The researchers’ work, anchored in the GKP encoding, provides a powerful lens through which to view the relationship between discrete and continuous quantum resources.

Resource Interplay Mirrors DV Magic, Entanglement, Coherence

The relationship between quantum resources driving computational power in continuous-variable and discrete-variable systems is now demonstrably parallel, according to new findings. The GKP encoding allows for a direct interpretation of computational complexity using the established discrete-variable resource-theoretic framework, but extending this understanding beyond GKP-encoded computations proved challenging until now.

While connections between continuous-variable non-Gaussianity and discrete-variable magic, and between mode-entanglement and logical qubit entanglement, were previously understood within the GKP encoding, a continuous-variable counterpart to discrete-variable coherence remained elusive. Symplectic coherence, the researchers demonstrate, fills this critical gap, completing the framework for understanding the resources enabling universal continuous-variable quantum computation.

The identification of symplectic coherence hinges on the Fourier transform (F) gate, a key component in continuous-variable computations. This gate’s role in inducing symplectic coherence establishes a clear link to the Hadamard gate in discrete-variable systems, which generates coherence through its operation on qubits. The team’s work reveals a practical implication for assessing the limits of classical simulation. The researchers’ analysis extends beyond simply identifying these resources.

They demonstrate how their interplay dictates the complexity of continuous-variable computations. The interplay of CV non-Gaussianity, entanglement, and symplectic coherence generalizes the resource interplay of DV magic, entanglement, and coherence in GKP-encoded quantum computations.

CV Resource Theory Identifies Advantages Over Classical Systems

The algorithm, based on quadrature propagation across multiple computational paths, mirrors recent simulation techniques used in discrete-variable systems but had remained largely unexplored in the continuous-variable regime. This parallel extends beyond mere functional similarity; the researchers demonstrate a direct correspondence between the resources themselves. The researchers write, highlighting the unifying nature of their discovery.

The significance of this correspondence lies in its potential to apply insights from the more established field of discrete-variable quantum computing to advance continuous-variable approaches. The researchers’ analysis focused on identifying the resources contributed by each quantum gate in universal gate sets across discrete-variable, GKP-encoded continuous-variable and general continuous-variable computations.

Their theorem formally establishes symplectic coherence as a computational resource, demonstrating that its absence allows for efficient classical simulation of quantum expectation values of constant-degree polynomial Hamiltonians. They state, solidifying the connection between the two computational paradigms.

👉 More information
🗞 Interplay of resources for universal continuous-variable quantum computing
✍️ Varun Upreti and Ulysse Chabaud
🧠 DOI: http://link.aps.org/doi/10.1103/2str-dhrg

More like this