State-of-the-art cosmological simulations on classical computers are limited
by time, energy, and memory usage. Quantum computers can perform some
calculations exponentially faster than classical computers, using exponentially
less energy and memory, and may enable extremely large simulations that
accurately capture the whole dynamic range of structure in the Universe within
statistically representative cosmic volumes. However, not all computational
tasks exhibit a `quantum advantage'. Quantum circuits act linearly on quantum
states, so nonlinearities (e.g. self-gravity in cosmological simulations) pose
a significant challenge. Here we outline one potential approach to overcome
this challenge and solve the (nonlinear) Schrodinger-Poisson equations for the
evolution of self-gravitating dark matter, based on a hybrid quantum-classical
variational algorithm framework (Lubasch 2020). We demonstrate the method with
a proof-of-concept mock quantum simulation, envisioning a future where quantum
computers will one day lead simulations of dark matter.