Asymmetric Interferometers Improve Quantum Information Processing Performance and Rates

The fidelity of quantum operations, a critical factor in the development of quantum technologies, is demonstrably affected by the architecture of the photonic circuits used to implement them. Researchers now demonstrate that strategically introducing asymmetry into photon loss within complex interferometers can, counterintuitively, improve operational accuracy. F. H. B. Somhorst and J. J. Renema, from the MESA+ Institute for Nanotechnology at the University of Twente, alongside colleagues, detail these findings in their article, ‘Mitigating quantum operation infidelity through engineering the distribution of photon losses’. Their work reveals a fundamental relationship between minimising errors and maximising the probability of a successful operation, challenging the conventional assumption that balanced loss is universally optimal for quantum information processing. An interferometer is a device that uses the interference of light waves to measure or manipulate them, and is a key component in many quantum technologies.

Photonic quantum information processing increasingly relies on integrated linear optical circuits, enabling the construction of compact and stable interferometric networks that precisely manipulate quantum states of light. These circuits facilitate the miniaturisation of components like beam splitters and phase shifters, crucial for building large-scale systems intended for quantum computation and communication. The ability to create complex quantum transformations stems from decomposing them into sequences of operations acting on pairs of optical modes, aligning naturally with the capabilities of these fundamental linear optical components.

Two prominent designs, triangular and rectangular multiport interferometers, have emerged as viable strategies for constructing these circuits, each possessing distinct characteristics regarding optical depth and loss. Recent research demonstrates these designs are interchangeable through systematic transformations, offering flexibility in circuit design and allowing researchers to tailor architectures to specific computational needs. Universal quantum information processing with linear optics necessitates projective measurements and feedforward mechanisms, introducing the nonlinearity required for creating entangling gates and conditional operations that overcome the limitations of purely linear systems. Operations are broadly categorised by the extent of measurement performed, ranging from full-measurement operations, such as boson sampling, to partial-measurement operations used for generating entangled states or performing photon distillation.

Circuit optimisation traditionally prioritises maximising the probability of successful operation under ideal conditions. However, practical imperfections, including photon loss, inaccuracies in transformation matrices, and errors in photonic indistinguishability, degrade the quality of feedforward information. Consequently, maximising success probability does not always translate to minimising quantum operation loss in real-world scenarios, demanding a more nuanced approach to design and optimisation. The rectangular multiport interferometer is often favoured for mitigating the effects of photon loss due to its inherent symmetry, preserving the target interference pattern subject only to uniform attenuation upon post-selection.

This work challenges the conventional assumption that balanced losses are always optimal for quantum information processing, exploring scenarios where asymmetrically designed interferometers may offer advantages for specific operations. The findings highlight a fundamental trade-off between minimising infidelity—the accuracy with which a quantum operation is performed—and achieving high success rates, crucial considerations for practical applications and the development of robust quantum systems. Researchers are meticulously examining how the architecture of these interferometers—specifically, the symmetry or asymmetry of photon loss within the network—impacts computational fidelity and efficiency.

Recent advances in photonic quantum computing increasingly focus on the precise construction and operation of multiport interferometers, complex networks that manipulate individual photons to perform computations. Researchers are actively investigating how the design of these interferometers—specifically, the symmetry of photon loss within the network—impacts the fidelity and efficiency of quantum operations. They demonstrate that asymmetric loss, deliberately introducing differing probabilities of photon loss at various points within the interferometer, can, in certain scenarios, improve the quality of quantum operations, stemming from the interplay between operation fidelity and success rate.

The study reveals a fundamental trade-off: minimising infidelity often necessitates accepting a lower success rate, and vice versa. Meticulously examining how asymmetric loss affects specific quantum operations identifies scenarios where deliberately uneven loss profiles enhance performance. This optimisation isn’t universally applicable; the benefits of asymmetry depend heavily on the specific operation being performed and the characteristics of the photonic qubits employed. The team utilises theoretical modelling and simulation to quantify the impact of different loss profiles, providing a detailed map of performance landscapes.

By carefully tailoring the asymmetry of loss, researchers can shift the balance between these two factors, optimising performance for particular tasks, and prioritising the successful transmission of photons along a specific path, even if it means accepting a slightly higher error rate on others. This is particularly relevant in applications where a successful outcome, even with some imperfections, is preferable to a failed attempt, highlighting the trade-off between minimising infidelity—the degree to which a quantum operation deviates from its intended behaviour—and achieving high success rates.

By embracing asymmetry, researchers can strategically sacrifice some degree of fidelity in exchange for a higher probability of success, paving the way for more robust and efficient photonic quantum computers and enabling practical applications in areas such as drug discovery, materials science, and financial modelling. This highlights the importance of nuanced design considerations in photonic quantum computing, demonstrating that simply striving for balanced loss isn’t always the most effective strategy. Instead, researchers must carefully tailor the loss profile of interferometers to the specific demands of the quantum operation, acknowledging the inherent trade-off between fidelity and success rate.

Quantum computation increasingly relies on multiport interferometers, complex networks of beam splitters and phase shifters manipulating single photons to perform calculations. Current efforts actively pursue multiple avenues, including boson sampling, linear optical quantum computing (LOQC), and more recent approaches like fusion-based computation, all aiming to realise scalable and fault-tolerant quantum processors. A significant emphasis lies on improving the performance of key components, with researchers continually refining single-photon sources and detectors, striving for higher efficiency and purity, recognising these as critical bottlenecks in photonic quantum systems. Parallel to this, substantial progress occurs in the development of integrated photonic circuits, which promise to miniaturise complex optical systems and enhance their stability and scalability, integrating multiple optical components onto a single chip, reducing both size and complexity.

The field actively explores the boundaries between quantum and classical computation, with studies investigating the limits of classical simulation of photonic linear optics, providing benchmarks for verifying the performance of quantum experiments and establishing demonstrable quantum advantage. Current efforts actively pursue multiple avenues, including boson sampling, linear optical quantum computing (LOQC), and more recent approaches like fusion-based computation, all aiming to realise scalable and fault-tolerant quantum processors. A significant emphasis lies on improving the performance of key components, with researchers continually refining single-photon sources and detectors, striving for higher efficiency and purity, recognising these as critical bottlenecks in photonic quantum systems.

Future work should prioritise addressing the scalability challenges inherent in photonic quantum computing, with integrated photonics offering a promising path towards miniaturisation. Further research is needed to increase the complexity and connectivity of these circuits. Simultaneously, investigations into novel error correction schemes and fault-tolerant architectures are crucial for building reliable and robust quantum processors. Exploring alternative photonic platforms and materials may also unlock new possibilities for enhancing performance and scalability. Continued investigation into the interplay between system design and operational performance remains vital, with a deeper understanding of the trade-offs between infidelity, success rate, and resource requirements enabling the development of more efficient and effective quantum algorithms and architectures. This requires a holistic approach, integrating theoretical modelling, experimental validation, and advanced characterisation techniques.