Periodic Reporting for period 2 - QUCHIP (Quantum Simulation on a Photonic Chip)
Periodo di rendicontazione: 2016-09-01 al 2018-02-28
QUCHIP aimed at implementing quantum simulation on integrated photonic processors. Photons present several unique advantages deriving from their mobility and the immunity to decoherence: these two features make them substantially different from any other quantum system. Furthermore, recent developments of integrated quantum photonics open novel prospectives towards reaching high dimensional systems. QUCHIP exploited these advantages to implement quantum walk experiments in which several photons propagate over complex circuit architectures “jumping” between different waveguides. This implements the most resource-efficient quantum computation scheme to date, known as Boson Sampling, which is one of the most promising to achieve the quantum supremacy regime in which quantum systems surpass classical ones. QUCHIP aimed at developing new photonic technologies, and at developing practical schemes to demonstrate key structural and functional elements of the network dynamics.
In addition, quantum technologies are also excellent candidates to investigate complex quantum phenomena, where classical approaches are doomed to failure due to the lack of computational power or sensitivity to the quantum mechanisms. The QUCHIP consortium has achieved significant results also in this direction by exploiting the properties of single photons and of effective integrated hardware to simulate the behavior of various relevant quantum processes.
Besides developing the technology to implement innovative quantum computation and simulation platforms, the consortium has also developed methods to verify their correct operation. Indeed, how can we be sure that a quantum computer is working properly if there is no way we can reproduce its outcomes with a classical (reliable) device? The QUCHIP project has made several seminal contributions to this fundamental issue. In fact, several methodologies have been developed, the most recent of which was based on the application of machine-learning techniques. The latter is a promising efficient approach, for instance, in experiments aimed at gaining insights into enormous amounts of (quantum) data without strong knowledge of the phenomenon.
The results of the QUCHIP project have been disseminated in 75 publications in high-impact scientific journals, and 268 presentations at national and international scientific conferences. During the project, QUCHIP partners have also disseminated their results to the general public using diverse communication strategies, namely, illustrative videos broadcast on YouTube and conferences open to the general public, as well as divulgation events directly involving young people at high schools.
As a novel perspective, the QUCHIP consortium has contributed to a rapidly emerging field that seeks to merge quantum computation and machine learning. The latter is a set of established computational methodologies in the classical domain that are particularly suitable for handling large sets of data, such as those produced by quantum devices. Merging of these two fields promises to open novel research lines that will lead to completely new approaches towards more efficient quantum technologies. In particular, machine learning techniques for benchmarking quantum technologies were applied for the first time in the QUCHIP project.
Notably, industrial applications can be foreseen for the hardware developed in the QUCHIP project (photon sources, integrated circuits and photon detection). Indeed, the number of enterprises developing quantum technologies is increasing worldwide and represent a potential market for the components and methodologies developed by the consortium. Furthermore, the scientific results obtained can open novel applications for Quantum Computing and Quantum simulation. More specifically, the hardware and theoretical tools obtained significantly enlarge the dimension and complexity of the quantum systems that can be investigated with a photonic platform.