29 Febbraio 2024

New advancements in quantum physics

The University of Trento is part of two international studies in the field of photonics. Mirko Lobino (Professor at the Department of Industrial Engineering) has fabricated a device that could reveal new properties of matter through laser beams. For the first time, computational chips use machine learning

Compared to traditional computers, quantum computers are like telescopes, they will make it possible to look farther. And, in particular, to observe what happens in the world of the ultra-tiny because they will be capable of operating and solving problems at an exponentially faster speed. Building these devices is the goal of those who study physics and quantum computing.
The University of Trento is taking part in this effort with Mirko Lobino, Professor of Physics at the Department of Industrial Engineering, who has recently moved his research activity on photonic chips from Australia to Trento. He is involved in two international projects led by his colleague Alberto Peruzzo, a Professor at the RMIT University.
The light that reveals matter. Research is opening the way to the development of a more compact and adaptable platform for classical and quantum photonic processors.
In this first study, the researchers built an integrated and reconfigurable photonic device, capable of reproducing, through light rays, environmental disturbances, external effects and interference. Quantum phenomena that occur within matter, which are difficult to observe and measure because they involve few electrons.
This device makes it possible to simulate the quantum properties of solid-state materials using light. It is made of lithium niobate, one of the most widely used materials for high-speed electro-optic modulators. Inside it, a series of tiny channels, a few micrometres long, direct light and in this way recreate the wave-like behaviour of electrons in an atomic lattice.
The device can be reconfigured multiple times, each time modelling a solid with different properties. For the first time, a single device was used, capable of simulating electronic transport across 2500 different configurations.
With this study, the researchers explain, different physical dynamics have been experimentally demonstrated on a single device, which is controllable, can be rapidly reconfigured with low power consumption and is small in size.
Machine learning for quantum computing. The second study for the first time uses machine learning to program reconfigurable photonic chips and control what happens inside these quantum devices.
The researchers simulated a small photonic quantum computer operating on a few quantum bits (the minimum amount of information in the quantum world).
However, with larger circuits, that use tens of thousands of qubits, things get complicated. Controlling these systems, just like understanding what happens inside a black box, is almost impossible. With black boxes, researchers know the inputs and outputs, how they respond to stimuli, but their internal mechanisms are completely unknown. The researchers would like to open this box and reveal its mechanisms. But how? "We will use machine learning techniques. We will use algorithms that can, on their own, measure and describe the process of light transformation.
The programming of the algorithms and the procedures for controlling the photonic chip were developed by the group of researchers led by Alberto Peruzzo.
UniTrento contributed to both projects. The group led by Professor Lobino manufactured the entire photonic device. He had worked on the creation of the device when he was at Griffith University. Once back in Italy, he continued his studies on quantum photonics in the new Integrated Quantum Photonics Laboratory of the University of Trento.
"We have built a photonic chip – he explains – that allows us to recreate more than two thousand different configurations of solid materials. Without having any of those materials. Instead of electrons we used laser beams. And we were able to see and measure the effects of what happens inside the circuit."
"One of the fundamental aspects of the quantum computer - he continues - is that it will allow us to efficiently simulate materials, optimize chemical reactions or discover properties that we may not yet know."
"These works could help us control very complex photonic devices, while keeping track of the physical processes that occur."
"Now we have simulated the known, theories that exist – he concludes – but we could try to engineer things that do not exist today or that we have not yet seen."

The first study, "Programmable high-dimensional Hamiltonian in a photonic waveguide array", has been published in Nature Communications and is available at this link: (DOI:
The second study, "Experimental graybox quantum system identification and control", has been published in Npj Quantum Information and can be found at this link: