Benchmarking quantum trial wavefunctions for phaseless auxiliary-field quantum Monte Carlo

The phaseless auxiliary-field quantum Monte Carlo (ph-AFQMC) method mitigates the fermionic sign problem by constraining stochastic propagation with a trial wavefunction, thereby introducing a controllable bias. However, classically constructing trial states that accurately capture many-body correlations remains a major challenge. In a pioneering advance, Huggins et al. proposed the quantum-computing-enhanced (QC-AFQMC) framework, in which correlated quantum trial wavefunctions are prepared for use in ph-AFQMC simulations. Subsequent work has improved the efficiency and viability of QC-AFQMC via matchgate-shadow overlap estimation and GPU-accelerated post-processing. Nonetheless, fundamental questions remain regarding which families of quantum trial states are most effective for QC-AFQMC. In this work, we present a comprehensive benchmarking study of quantum trial wavefunctions, encompassing the Unitary Coupled Cluster (UCC) family, hardware-efficient, adaptive, and other variational ansatze, to evaluate their accuracy, expressibility, and scalability within the QC-AFQMC framework. We test these ansatze on molecular systems with high qubit counts under bond stretching and demonstrate large-scale parallelization of the trial-state optimization across multiple GPUs and compute nodes.