Harnessing Noise as a Resource in Quantum Algorithms for Open Quantum System Simulations and Gibbs State Preparations

Noise is widely viewed as the central barrier to scalable quantum computation, motivating extensive efforts toward suppression, correction, and fault tolerance. Yet for many tasks, it can be used as a resource. In this talk, I will delineate how noise can be harnessed as a computational resource for developing quantum algorithms tailored to near-term, noisy hardware. I will first demonstrate a noise-assisted algorithm that selectively preserves physical noise to emulate non-unitary channels of open quantum systems. By partially encoding environment channels and partially error correcting the excessively encoded channel, our algorithm reduces the need for ancilla qubits. Complementing this, I will also demonstrate that noise can actively accelerate Gibbs state preparation in many-body systems. Drawing on the Eigenstate Thermalization Hypothesis, I will show that controlled noise enhances scrambling, promotes ergodicity, and shortens equilibration times. Interestingly, I will show that such an advantage can be derived merely by phase-flip channels in experiments. Together, these results establish a new paradigm: rather than eliminating noise at all costs, one can co-design algorithms with noise to drive useful physical processes. This perspective opens promising routes for practical open-system simulation, state preparation, and dynamical tasks on current and future quantum devices, reframing noise from an adversary into a tool for deriving advantages before fault-tolerance.