Dipartimento di Fisica

Quantum Lab group Sapienza Università di Roma (Rome) recently published two groundbreaking studies in international journals where they describe a new set-up for a quantum-to-quantum Bernoulli factory, one in Nature Photonics in collaboration with INL (International Iberian Nanotechnology Labs) and IFN-CNR (Institute for Photonics and Nanotechnologies-National Research Council), and, a second one, in Science Advances in collaboration with INL.

A Bernoulli factory is a technique for manipulating randomness, utilizing random coin flips with a specified probability distribution as inputs, and producing coin flips with a different, desired distribution as outputs. Consider a scenario where we have a coin that lands on heads with an unknown probability. Our goal is to create a new coin that lands on heads with a different probability, potentially defined by a function of the initial probability. The Bernoulli factory cleverly allows us to flip the original coin multiple times and leverage the various outcomes to simulate a new coin with the desired probability. This type of randomness manipulation serves as a crucial subroutine in various probabilistic computations, such as numerical integration and Monte Carlo methods for simulating physical systems. In the quantum mechanics framework, these protocols follow a similar principle, with the distributions encoded as quantum states at both the input and output, and so the name “quantum-to-quantum Bernoulli factory”.

The unique features of Bernoulli factories have inspired collaboration among three research teams to explore various methods for implementing these factories using single-photon states, linear optical circuits, and photodetectors tailored for specific applications. At the heart of the photonic quantum-to-quantum Bernoulli factory lies the optical interferometer, which serves as the fundamental building block of a photonic quantum information processor. In this setup, the dynamics are entirely governed by the configurations of linear optical elements, allowing for the manipulation of both the path and polarization degrees of freedom of photons. Given the statistical properties that photons exhibit, their evolution within these interferometers becomes computationally challenging for classical computers to simulate

In this context, we have developed two platforms that manipulate distinct degrees of freedom of single-photon states. One implementation [1], in collaboration with INL and IFN-CNR, works with the so-called path-encoded qubits, where the information is written in the path of each photon. Such an encoding is particularly suited for integrated optical circuits. This quantum version of Bernoulli factory was realized in a 6-mode universal and programmable photonic chip, fabricated by IFN-CNR. In this circuit, the single-photon qubits are processed through waveguides, integrated beamsplitters, and phase shifters. Such architectures are the ideal solutions to use the Bernoulli Factory in the quantum computing framework, for example, as a subroutine of quantum photonic hardware

In the second platform [2], developed in collaboration with INL, where the qubits are encrypted in the polarization states of single photons, we used a quantum dot-based single-photon source that is at the forefront of photonic quantum technologies. Also in this implementation, we demonstrated all the necessary dynamical steps required to realize arbitrary quantum Bernoulli factories. The main advantage of this second architecture is that everything was realized through in-fiber and bulk optics elements, which is particularly suitable to interface quantum Bernoulli factory with quantum networks, and in general, enables the possibility of adding this functionality within complex quantum communication and cryptographic protocols. This second work was published in Science Advances.

These advancements represent significant steps in exploring the information-processing capabilities of quantum light. Bernoulli factories have emerged as a promising framework in which quantum machines exhibit distinct advantages over their classical counterparts. By leveraging the unique properties of quantum light, researchers can unlock new possibilities for efficient computation and sophisticated randomness manipulation, paving the way for innovative applications in various fields, from cryptography to computing and simulation.

The two works are parts of the European projects “PHOQUSING – PHotonic QUantum SamplING Machine” and “QU-BOSS” both coordinated by Prof. Fabio Sciarrino, group leader of Quantum Lab at Physics Department of Sapienza University of Rome, and the by Italian project ICSC– Centro Nazionale di Ricerca in High Performance Computing, Big Data and Quantum Computing.

References

[1] Francesco Hoch et al., Modular Quantum-to-Quantum Bernoulli Factory in an Integrated Photonic Processor, Nature Photonics (2024) https://www.nature.com/articles/s41566-024-01526-8
[2] Giovanni Rodari et al, Polarization-encoded photonic quantum-to-quantum Bernoulli factory based on a quantum dot source. Science Advances 10,eado6244(2024). DOI:10.1126/sciadv.ado6244 https://www.science.org/doi/full/10.1126/sciadv.ado6244

Press Release:

Comunicato ANSA: https://www.ansa.it/amp/canale_scienza/notizie/frontiere/2024/10/08/la-luce-si-prepara-a-rivoluzionare-i-computer_acdb5b13-11ec-4e55-8f31-a4670f6cbf07.html
Comunicato ICSC: https://www.supercomputing-icsc.it/2024/10/10/un-nuovo-approccio-fotonico-alla-manipolazione-della-casualita-nei-computer-quantistici/

Sito Quantum Lab: https://www.quantumlab.it/

Authors: Francesco Hoch, Giovanni Rodari, Taira Giordani, Alessia Suprano, Luca Castello, Elena Negro, Gonzalo Carvacho, Nicolò Spagnolo, Francesco Ceccarelli, Ciro Pentangelo, Simone Piacentini, Andrea Crespi, Roberto Osellame, Ernesto F. Galvão, and Fabio Sciarrino