A new benchmark tests quantum algorithms with realistic noise

Researchers have established a new benchmark for evaluating quantum algorithms by testing their ability to prepare ground states amidst realistic noise. The work, centered on quadratic fermionic Hamiltonians exhibiting two distinct phases separated by a quantum phase transition, allows for both analytical and numerical comparisons of algorithms, a rare combination for assessing performance. Results show multi-frequency cooling algorithms, as proposed in Phys. 7 (2025) 033162, can outperform adiabatic evolution in the topological phase, where gap-closing typically limits adiabatic protocols. The team further demonstrated enhanced robustness to parameter imperfections in the cooling protocol, suggesting a practical advantage for noisy quantum state preparation. An exactly solvable family of quadratic fermionic Hamiltonians is now a benchmark for evaluating quantum algorithms under realistic conditions, allowing for both analytical and numerical validation of performance. This model facilitates a direct comparison…