Quantum Device Benchmarking from Many-Body Quantum Chaos

Here we present a simple and efficient benchmarking protocol to estimate the many-body fidelity between a target state and the actual state obtained from experimental evolution. Our protocol relies only on time evolution of a quantum system undergoing chaotic dynamics, followed by projective measurements in a fixed local basis without any local control; this is in stark contrast to many existing methods which require fine-tuned spatiotemporal control and substantial experimental resources that scale exponentially with system size. Fundamentally, this simplification stems from a universal phenomenon in many-body quantum chaos: the emergence of universal random statistics in a local region. We demonstrate our benchmarking protocol numerically for random unitary circuits, and experimentally using a Rydberg quantum simulator.