TL;DR: Our results provide strong evidence that the quantum method gets strong performance with fewer than 60 logical qubits and shows four to six orders of magnitude smaller machine size than the classical and QRAM-style baselines on the main real-world datasets. Rather than fearing that classical