Swinburne University of Technology Team Introduces Matrix Phase-Space for Quantum Simulation

Researchers at Swinburne University of Technology have introduced a novel phase-space representation that incorporates global quantum symmetries, significantly advancing the field of many-body quantum simulations. Peter D. Drummond and colleagues present a ‘matrix phase-space’ method which projects calculations onto a reduced Hilbert space, thereby minimising the statistical errors inherent in these complex simulations. This unified approach not only builds upon, but also improves existing phase-space techniques, demonstrated through a detailed application to Gaussian boson sampling. Exploitation of parity symmetry, a fundamental property of quantum systems, sharply reduces errors compared with previous methodologies.

Matrix phase-space streamlines error reduction in Gaussian boson sampling simulations

Sampling errors in Gaussian boson sampling were reduced by very large factors, exceeding improvements observed with previous techniques, enabling validation of outputs with a level of accuracy previously unattainable. The matrix phase-space approach, developed at Swinburne University of Technology, projects calculations onto a reduced Hilbert space, effectively diminishing the computational burden. This streamlined process allows for more efficient simulation of complex quantum systems, representing a substantial leap forward in computational power and opening avenues for exploring larger and more intricate quantum phenomena. The significance lies in the ability to accurately model systems that were previously intractable due to the exponential growth of computational requirements with system size.

It unifies several existing phase-space methods, including the stochastic Bloch and gauge-P representations, providing a single, coherent framework for analysing quantum behaviours. These methods, while useful individually, often suffer from limitations in accuracy or applicability to certain types of quantum systems. The matrix phase-space method addresses these shortcomings by providing a more general and robust approach. Substantial reductions in sampling errors during Gaussian boson sampling (GBS) verification constitute a key advancement in simulating quantum systems, achieved at Swinburne University of Technology. GBS, a promising candidate for demonstrating quantum supremacy, relies heavily on accurate simulation to validate experimental results and benchmark performance. The matrix phase-space method utilises global quantum symmetries to project calculations onto a smaller, more manageable computational space. This directly addresses the issue of distribution tails that plague traditional phase-space methods, particularly in low-loss, number-resolved scenarios where accuracy diminishes rapidly. These tails represent rare but significant events that contribute disproportionately to the overall error. The technique mitigates these inaccuracies, improving upon methods like the Wigner and Q-function which exhibit exponentially growing errors as the system complexity increases, and its ability to handle both conserved and partially conserved symmetries expands its applicability beyond idealised systems, allowing for more realistic modelling of physical processes. Conserved symmetries, such as energy and momentum, are fundamental properties of physical systems, while partially conserved symmetries arise in systems with dissipation or decoherence.

Rigorous mathematical foundations underpin improved quantum simulation and verification techniques

The development of matrix phase-space offers a promising route to tackling the computational bottlenecks inherent in simulating many-body quantum systems, particularly as scientists strive to build and verify increasingly complex quantum computers. The exponential scaling of Hilbert space dimension with the number of quantum particles presents a formidable challenge to classical simulation. By projecting onto a reduced Hilbert space, the matrix phase-space method effectively reduces the dimensionality of the problem, making it tractable for classical computers. The team at Swinburne University of Technology explicitly frames this work as a methodological advance, prioritising detailed proofs of basic theorems and operator identities over a broad exploration of applications. This emphasis on mathematical rigour ensures the validity and reliability of the method, providing a solid foundation for future research and development. The detailed proofs establish the theoretical underpinnings of the method, while the operator identities provide a set of rules for manipulating quantum operators within the matrix phase-space framework.

A reduction in errors for Gaussian boson sampling is a concrete example of the benefits as scientists build larger quantum computers and need ways to verify their outputs. Verifying the outputs of quantum computers is crucial for ensuring their correctness and reliability. As quantum computers scale up in size and complexity, the task of verification becomes increasingly challenging. The matrix phase-space method provides a powerful tool for addressing this challenge. Swinburne University of Technology established a new computational framework, matrix phase-space, designed to incorporate global quantum symmetries directly into phase-space simulations. Phase-space methods represent quantum states using classical variables, allowing for the application of classical computational techniques. Incorporating symmetries into the phase-space representation further enhances the efficiency and accuracy of the simulation. Calculations are projected onto a reduced Hilbert space, a mathematical space defining all possible quantum states, thereby streamlining complex simulations and potentially minimising errors. This projection is achieved through a carefully designed operator that preserves the essential quantum properties of the system. By unifying existing phase-space techniques like the stochastic Bloch and gauge-P representations, a single, more flexible approach to modelling quantum systems has been created. This unification simplifies the development of new quantum algorithms and allows for the seamless integration of different simulation techniques. The method’s ability to accurately simulate quantum systems with 01% error represents a significant improvement over existing techniques, paving the way for more reliable quantum computations.

The researchers developed a new computational framework, matrix phase-space, which incorporates global quantum symmetries into simulations. This method projects calculations onto a reduced Hilbert space, thereby improving the efficiency and accuracy of many-body quantum simulations and unifying existing phase-space techniques. Applying parity symmetry demonstrably reduces sampling errors, achieving results with 0.1% error in tests of Gaussian boson sampling. The authors detail the theoretical foundations and operator identities of this approach, providing a basis for future work in verifying outputs from quantum computers.