Single Chip Now Performs Multiple Quantum Calculations Efficiently

Scientists at Warsaw University, led by Jacek Gosciniak, have developed a recirculating bricks mesh architecture capable of performing multiple functions within a single optical system. This technology expands the range of applications to include boson sampling, a computational task frequently used to demonstrate quantum advantage, and the crucial determination of photon indistinguishability. The system exhibits flexibility with respect to both spatial and temporal modes of light, thereby broadening its utility in the field of photonic quantum technologies.

Reduced photonic loss enables scalable boson sampling and quantum computation

Optical losses within photonic circuits represent a significant impediment to the development of complex quantum systems. The research team has successfully reduced these losses to between two and four Mach-Zehnder interferometers (MZIs) per unit cell, a substantial decrease compared to the six previously required in state-of-the-art hexagonal meshes. MZIs are fundamental building blocks in integrated photonics, functioning as beam splitters and phase shifters to manipulate light paths. Maintaining signal integrity becomes increasingly difficult with a higher component count, as each MZI introduces a small amount of loss; however, this reduction in loss unlocks the potential for scaling up photonic quantum systems to more complex and powerful configurations. The recirculating “bricks” mesh architecture, demonstrated by Jacek Gosciniak and colleagues, enables both boson sampling, a key demonstration of quantum advantage over classical computers, and precise determination of photon indistinguishability, a critical requirement for reliable quantum computation. Boson sampling involves calculating the probability distribution of indistinguishable photons emerging from a complex network, a task that becomes exponentially harder for classical computers as the number of photons increases.

A recirculating “bricks” mesh architecture underpins the system, enabling various functions within a single chip. Unlike traditional feed-forward networks where the size of the circuit grows linearly with complexity, this recirculating design allows for scaling both the number of modes and photons used in experiments without a corresponding increase in the physical size of the components. This is achieved through the clever reuse of optical pathways, reducing the overall footprint and minimising signal degradation. Current results focus on boson sampling and determining photon indistinguishability, key elements in photonic quantum technologies, under controlled laboratory conditions. The ability to accurately characterise photon indistinguishability is paramount, as it ensures that photons behave as truly identical particles, which is essential for performing interference-based quantum computations. Minimising the component footprint also reduces optical propagation losses, preserving the delicate quantum properties of the light signals, such as superposition and entanglement, and further work will focus on achieving performance improvements and scalability for practical applications. The reduction in loss is achieved through careful optimisation of the waveguide geometry and material selection, minimising scattering and absorption of light.

Integrated boson sampling and indistinguishability testing via a reprogrammable photonic processor

Photonic quantum computing holds the promise of significant speed advantages for specific computational problems, such as factoring large numbers or simulating quantum systems, yet building reliable and scalable systems remains a formidable challenge. Programmable optical processors are now being demonstrated with increasing sophistication, aiming to integrate multiple quantum functions onto a single chip, offering a versatile platform for exploring different quantum algorithms. Current designs often rely on complex two-dimensional meshes of Mach-Zehnder interferometers, components that split and redirect light, and scaling these up without significant signal loss is proving difficult. The challenge lies in maintaining coherence, the preservation of the quantum phase of light, as it propagates through the numerous optical elements.

Extending functionality to temporal modes alongside spatial modes represents a key step towards flexible photonic quantum processors. Spatial modes refer to the shape of the light beam, while temporal modes describe the structure of the light pulse in time. By exploiting both spatial and temporal degrees of freedom, the information capacity of the system can be significantly increased. Adjustable loop implementation within the mesh achieves this broadened potential, offering a flexible platform for future quantum technologies. This is accomplished by incorporating tunable delays and phase shifters into the recirculating loops, allowing for precise control over the temporal properties of the light. By demonstrating versatility across both spatial and temporal light modes, the architecture has created a more adaptable platform for diverse quantum technologies, potentially enabling the implementation of more complex quantum algorithms and the exploration of new quantum phenomena. The ability to programmatically control these modes is crucial for adapting the processor to different tasks and optimising its performance. The system’s reprogrammability is achieved through the use of electronically controlled phase shifters, allowing for dynamic reconfiguration of the optical circuit.

The researchers demonstrated a single programmable optical system capable of performing boson sampling and assessing photon indistinguishability. This versatility is achieved through a two-dimensional hexagonal mesh of Mach-Zehnder interferometers, extending functionality to include both spatial and temporal modes of light. Exploiting these temporal modes alongside spatial modes increases the information capacity of the system and offers a flexible platform for photonic quantum technologies. The authors suggest this architecture provides a means of implementing more complex quantum algorithms and exploring new quantum phenomena.