Error Mitigation for Quantum Simulations Beyond Classical Limits

As quantum processors approach regimes that challenge the best available classical methods, the ability to handle errors becomes critical for unlocking their full potential. Error mitigation is essential for obtaining accurate results at large scales on both contemporary and error-corrected devices. 

In this talk, I will present recent theoretical and experimental advances demonstrating how effective error mitigation can push quantum simulations on leading hardware platforms beyond the limits of classical approaches. I will highlight results obtained by Qedma's characterization-based mitigation software QESEM on IBM’s Heron processors. I will show how QESEM achieves accurate expectation values of observables in large-scale quantum simulations of spin models in regimes that challenge the capabilities of state-of-the-art tensor-network and Pauli-propagation methods on modern HPC systems. I will also describe quantum simulations on Quantinuum systems that use Qedma's low-overhead, circuit-agnostic mitigation to obtain high-accuracy results at classically challenging parameter regimes. In addition, I will discuss how the utility of quantum computers can be extended by combining QESEM with HPC systems. I will present how QESEM can provide unbiased estimators for tensor-network backpropagated observables, and show demonstrations utilizing this approach to execute significantly deeper circuits on current hardware.

Together with insights from recent theoretical work, these results highlight the essential role of error mitigation in enabling practical quantum advantage as quantum hardware transitions from the NISQ era toward fault tolerance.