Noise-resilient Gibbs sampling protocols with temperature scaling guarantees

Construct Gibbs sampling protocols that provably enable efficient thermal state preparation at temperatures proportional to the hardware noise level, validating applicability in late NISQ and early fault-tolerant regimes.

Background

The authors speculate that certain Gibbs sampler variants could prepare thermal states efficiently at temperatures set by hardware noise, potentially making them suitable for late NISQ and early fault-tolerant devices.

They note that no protocols currently provide such guarantees, although related noise resilience has been proven for a dissipative ground-state preparation algorithm, motivating the search for new quantum algorithms with natural error resilience.