Quantum system of nine atoms outperforms networks with thousands of nodes

Researchers built a quantum reservoir using nine interacting atomic spins controlled by NMR.

They deliberately used dissipation to provide an optimal short-term memory for prediction tasks.

On NARMA benchmarks and multi-day temperature forecasts, the nine-spin system beat classical echo state networks scaled to thousands of nodes.

The study was published in Physical Review Letters and shows small, noisy quantum systems can solve certain time-series problems.

The approach is specialized and not a general-purpose quantum computer, and scaling remains a challenge.