Noise Helps Quantum Computers ⚡ экспресс
In the era of noisy quantum devices, noise is typically seen as a hindrance, but some processes can harness it. Based on the eigenstate thermalization hypothesis (ETH), it is shown that noise accelerates the preparation of Gibbs states. The effect is reproduced on a spin-1/2 chain model with a local Hamiltonian: in the non-integrable case, thermalization is sped up by ~3.5 times when noise (Haar or phase) is added, while an integrable system, which normally does not thermalize, reaches a thermal state under the influence of noise. Since verifying a local Gibbs state on a quantum processor is relatively easy, the proposed approach offers a practical solution to an important problem. The findings establish a new paradigm: noise can be purposefully exploited to gain benefits on quantum computers long before full fault tolerance is achieved.
To quickly dissolve sugar, you don't need to stir carefully — just shake the glass. A similar principle works in quantum systems: noise, usually considered a disturbance, speeds up the transition to a thermal state. This is a state of complete chaos, where everything is mixed to the limit from the point of view of entropy — a measure of disorder.
Previously, noise was thought to interfere. But it turns out that without it, some systems cannot reach equilibrium at all.
Properly dosed noise works like a shake — helping the system shuffle everything faster.
This discovery is important for practice. Thermal states are needed for calculations, for example, to model materials. They are easy to verify, which means that today we can extract benefit from imperfect quantum machines. Paradoxically, the noise that engineers try to eliminate has turned out to be the key to the first useful quantum computations.
🎯 In some quantum models, without noise, the system freezes forever in its initial state — adding noise literally "turns on" its evolution.