LSU Built a 40-Photon Quantum Machine Out of Ordinary Laser Light. It Runs at Room Temperature and Learned Six Functions Without Being Rewired.

Quantum computers are supposed to solve problems that overwhelm ordinary machines, but the quantum states they run on are so fragile that a stray vibration or a lost photon can wreck a calculation. Physicists at Louisiana State University have now shown a way around that fragility by starting from the opposite end. Rather than manufacturing delicate quantum light, they began with bright, ordinary classical laser light, passed it through an optical network, and then counted the photons arriving at the far end one by one. That counting step, they report in the journal Advanced Science, reveals multiphoton quantum behavior that was hiding inside the classical beam all along, and it can be put to work processing information.

The trick exploits a fact that is usually treated as a nuisance. A bright classical beam contains enormous numbers of photons, but the exact number hitting a detector fluctuates from one measurement to the next. Using photon-number-resolving detectors, which can tell whether one, two, ten or forty photons arrived at once, the team led by associate professor Omar S. Magaña-Loaiza selected specific photon-number events after the fact. Each such selection corresponds to a different multiphoton quantum system, and all of them live inside the same classical light field. By combining photon number with the light's polarization and its spatial structure, the researchers built what they call a multiphoton quantum reservoir: a vast web of possible states and connections that, in the experiment, provided 861 measurable components. The platform reached systems of up to 40 photons, operated at room temperature, and kept working in the presence of substantial noise and loss.

Reservoir computing is a way of using a physical system's natural complexity to do hard calculations without controlling every step. Information goes in, ripples through many interconnected paths, and produces a rich pattern at the output; only a simple readout has to be trained to interpret that pattern. In the LSU device, the light itself performed the complicated transformations, and the photon counts revealed the result. "Rather than requiring perfectly isolated and extremely fragile quantum systems, we show that useful quantum behavior can be extracted from ordinary classical light, even in the presence of substantial noise and loss," Magaña-Loaiza said. "Our system operates at room temperature and gives us access to multiparticle quantum systems containing up to 40 particles."

The team first used the platform as a quantum simulator. Photon-number measurements reproduced the characteristic spreading of a quantum random walk even when the optical network was noisy, and a synthetic lattice built from different states of light reproduced thermalization and anti-thermalization, the processes by which fluctuations in a many-particle system grow or shrink over time. Then they asked whether the same device could learn. With the optical network frozen in a single randomly chosen configuration and inputs encoded in the light's polarization, the reservoir turned each input into a far richer pattern across spatial modes and photon numbers. Training only the final readout, the system learned six very different mathematical functions without any physical reconfiguration. For the most nonlinear tasks, using the complete photon-number distribution improved the predictions, evidence that higher-order multiphoton correlations are doing real computational work.

"We combined the light's polarization and spatial structure with photon-number-resolved measurements in a single system," said Mingyuan Hong, the study's first and corresponding author. "This creates a much richer information space. The same optical platform can explore multiparticle quantum dynamics and learn mathematical functions with very different behavior." The collaboration included researchers from the Universidad Nacional Autónoma de México and the Universidad Politécnica de Pachuca.

The significance is less about any single benchmark than about where the quantum resource comes from. Bright classical light is cheap and abundant; genuine multiphoton quantum states are neither. If distinctly quantum measurements can pull usable multiparticle behavior out of a classical source in a warm, noisy room, that points toward quantum technologies built for real laboratories rather than perfectly isolated ones. The authors note that scaling to larger photon numbers will depend partly on faster and more capable photon-counting detectors, but the approach, they write, offers a route to bigger multiphoton systems using tools that already exist.