Researchers advance quantum simulation error limits and network optimization

Researchers have developed new methods to address computational challenges in quantum simulation and large-scale network optimization.

A team involving the Technical University of Munich, the University of Innsbruck, and the Austrian Academy of Sciences has demonstrated a way to experimentally characterize quantum simulators. As quantum systems become more powerful, verifying their results becomes difficult because they can no longer be easily replicated by classical computers. The researchers developed an approach that uses experimental data to learn how a simulator actually behaves, allowing them to translate uncertainties and noise into quantitative error limits. This method was tested using an ion-trap quantum simulator with up to 51 ions.

Separately, computer scientists at ETH Zurich have developed a new software program designed to solve massive network and optimization problems more efficiently. By utilizing Laplace equations and incorporating randomness to limit computational effort, the software overcomes the speed and memory limitations faced by previous methods. The research group has integrated these findings into the open-source software 'apxchol', which aims to accelerate complex calculations on supercomputers. These mathematical models are applicable to various fields, including power grids, traffic routing, and task distribution in artificial intelligence.