A recent experiment has shown that a many-body system of ultracold atoms can be used as a quantum simulator for experiments where classical computers fail. This also allows physicists to have a better understanding of how particles tunnel, and it opens up new avenues of study in condensed matter physics. A group led by Immanuel Bloch of Max Planck Institute of Quantum Physics and of Ludwig Maximilians University Munich has demonstrated that a quantum system can outperform classical calculations by confining a gas of supercooled rubidium atoms to an optical lattice and following the relaxation behavior of the system on a much larger timescale than any classical method could handle. An optical lattice formed by counterpropagating laser beams created a spatially periodic polarization pattern through the interference of the beams. The rubidium atoms were trapped in the dark and light areas of the lattice and became aligned in a regular pattern. The atoms in the lattice were then grouped in pairs into an optical superlattice by adding another light field with twice the spatial period of the original lattice. This created a density wave state far from the system’s equilibrium point. The atoms were allowed to tunnel along the spatial direction of the superlattice and to collide with one another on their way back to thermal equilibrium, creating complicated many-body dynamics. (a) Schematic shows how the atoms in an optical lattice relax from an excited density wave to a quasi-steady state. (b) The experimental data (blue circles) is very much in line with the simulation’s data (black line). However, the experiment could track the system’s behavior for a much longer period of time. J = strength of tunnel coupling, t = time. Courtesy of Max Planck Institute of Quantum Physics. After the system returned to equilibrium, as with a plucked string returning to rest, the system’s local density, tunnel currents and observable properties were probed for a variety of lattice heights and evolution times. They showed a rapid relaxation back to quasi-steady-state values and were in excellent agreement with previously computed numerical simulations. Classical computers can track many-body dynamics for a short period, and this provided a benchmark for the quantum experiment. The timescale of the experiment was much greater than that of the classical simulator’s predictions and could track the evolution of the system for far longer, giving much more precise data for a longer period of evolution.