A First: 3D-Chip-Based Analog Photonic Quantum Computer Demonstrates Quantum Fast Hitting

Press Release provided by
Xianmin Jin Research Team, Shanghai Jiao Tong University

Analog quantum computing has been an appealing tool with potential real-life applications on various optimization and simulation tasks, and much less stringent requirements on error corrections compared to universal quantum computing. Quantum walks, a key protocol for analog quantum computing, has theoretically shown many quantum advantages — for example in the speedup of fast hitting tasks in glued tree structures. However, in order to bring the advantages into reality, there exist many prerequisites on the physics quantum system. One is to make the system scalable so that it can cope with real problems of certain complexity.

The demonstration of analog quantum computing has normally been on a very small scale for proof-of-principle studies, and the number of photons is the main resource for expanding the quantum system. Until very recently, as Nature Photonics reported, a scalable analog quantum computing device was realized on an integrated quantum photonic chip, which opened a new roadmap for using the dimension and scale of the quantum evolution system as the new resource for analog quantum computing.

Schematic diagram of scalable analog quantum computing device for fast hitting tasks on hexagonal graphs.

In research conducted by Prof. Xian-Min Jin and his team in Shanghai Jiao Tong University, collaborating with Prof. Myungshik Kim in Imperial College London and Dr. Carlo Di Franco in Nanyang Technological University, hexagonal graphs as an alternative structure was proposed for fast hitting tasks instead of the traditional binary glued trees. The hexagonal graphs resemble the gluing of two tree-like structures, and are highly scalable to be mapped in the three-dimensional integrated quantum photonic chips, where the longitudinal direction represents the evolution time, and the cross-section view shows the hexagonal structure formed by waveguide arrays. The team led by Xian-Min Jin drew on their experimental expertise to report the largest-scale integrated photonic chip to demonstrate the first quantum walk experiment in real two-dimensional space.

For the research on quantum fast hitting, they injected photons from the “Entry” waveguide and measured the light intensity at the “Exit” waveguide as the “hitting efficiency.” Researchers found that quantum walks perform quadratically faster than classical random walks, showing the first experimental demonstration of quantum advantages in fast hitting tasks.

Schematic diagram mapping the hexagonal graphs for fast hitting onto the photonic chip.

A comparison of classical fast hitting and quantum fast hitting.

Experimental patterns showing the optimal quantum fast hitting on hexagonal graphs of different layer depths. In these graphs of different complexities, quantum fast hitting always has high light intensity in the “Exit” (the right-most waveguide).

The experimental implementation for fast hitting on glued trees represents exciting progress, as it realizes one of the most representative examples that quantum theoreticians have raised to showcase the speedup by quantum walks. Besides, considering the protocol’s essence as an optimization process and the similarity between binary trees and decision trees in computer science, the protocol of fast hitting on glued trees may further trigger many useful applications that utilize quantum speed-up for tasks such as logistics, finance and information searching. “Our demonstration can also be seen as a first step towards the realization of scalable quantum fast search. By adding more photons we can also build a large network of nonclassical states which can be used for applications as well as fundamental studies.” said Prof. Myungshik Kim.

The scalable quantum device enabled by the integrated quantum photonic chip also provides an excellent platform for quantum simulation of other physical systems, and may potentially stimulate research on many multidisciplinary topics and emerging scientific questions, including astronomy simulation, quantum machine learning, quantum topological photonics, quantum imaging for biological and medical applications, and so on.

This October the team led by Prof. Xian-Min Jin launched the first software for photonic analog quantum computing, FeynmanPAQS, with the aim of encouraging wider participation and brainstorming for various simulation proposals and potential applications connecting real problems. It would be delightful in the near future to see more quantum benefits being brought into reality.