T. Farajollahpour
Abstract
Realizing the promise of quantum computation for condensed matter many-body problems depends as much on software as on hardware, yet the area is reviewed far more often than it is quantified. We address this gap by pairing a focused survey of quantum algorithm software for condensed matter physics with a compact, fully reproducible benchmark suite that turns qualitative claims into concrete numbers. Each algorithm family, namely the variational quantum eigensolver (VQE), quantum phase estimation (QPE), quantum annealing and the quantum approximate optimization algorithm (QAOA), and quantum machine learning (QML), is demonstrated on a canonical lattice model and validated against an independent classical reference, from exact diagonalization and the Bethe ansatz to matrix-product-state DMRG. Within this suite we quantify two issues usually treated only qualitatively. Mapping the Fermi-Hubbard model to qubits under the Jordan-Wigner and Bravyi-Kitaev encodings, we tabulate qubit counts, operator weights, and gate costs and expose a geometry-dependent trade-off between the two. Simulating the circuits under a depolarizing noise model, we show that zero-noise extrapolation restores ground-state energies and optimization quality across the noise range. Around these results we review the algorithms as applied to strongly correlated systems, topological phases, and quantum magnetism, together with the leading software development kits (Qiskit, Cirq, PennyLane, and Q#) and the classical and tensor-network methods against which quantum approaches must be benchmarked. All circuits, seeds, and data are released so the benchmarks can be reproduced and extended. We argue that standardized, reproducible benchmarks of this kind are essential to gauge progress and identify genuine quantum advantage in condensed matter physics.