Quantum thermal state preparation algorithms and aspects of quantum advantage

We will talk about recent developments in exploiting thermalization in Nature as a quantum algorithmic tool to prepare thermal state and, consequently, the aspects of quantum advantage. First, we discuss how thermodynamic Lindbladians (Davies generators and their variants) inspire efficient quantum Gibbs sampling algorithms (2303.18224, 2311.09207). Next, we discuss two recent theoretical arguments for quantum advantage in this context: low-temperature cooling is universal for quantum computation (2309.16596), and constant-temperature Gibbs sampling is as hard as random circuit sampling (2404.14639).