Kinetic frustration enables single-molecule computation

A fundamental challenge in physical systems is implementing computation at the microscopic scale where thermal fluctuations dominate. While biological systems achieve this through complex molecular networks, the physical principles enabling simpler systems to process temporal information remain unclear. Here we demonstrate how non-equilibrium dynamics can enable single molecules to perform sophisticated computation through thermal-kinetic frustration-a principle that creates a controlled discrepancy between thermodynamic stability and kinetic accessibility. By engineering this frustration in a linear polymer with N binary-state units, we create a physical realization of a deterministic finite automaton capable of accessing 2N configurations through non-equilibrium driving, far exceeding the N + 1 configurations available at equilibrium. Despite operating in a thermal environment, the molecule's dominant configuration evolves deterministically, enabling recognition of complex temporal patterns through mechanical control signals. Our framework establishes how stochastic microscopic dynamics can give rise to deterministic computation, providing new insights into non-equilibrium statistical mechanics and information processing in physical systems. The theoretical predictions can be tested using DNA nanotechnology, with potential applications in biosensing and adaptive materials.