Quantum Flow algorithm: quantum simulations of chemical systems using reduced quantum resources and constant depth quantum circuits

TL;DR

This paper evaluates the Quantum Flow algorithm's ability to simulate chemical systems efficiently using fewer quantum resources and constant-depth circuits, demonstrating promising results with classical and quantum hybrid strategies.

Contribution

It introduces a quantum simulation approach combining classical downfolding with quantum flow optimization, reducing qubit requirements while maintaining accuracy.

Findings

QFlow-SD closely matches canonical UCCSD energies with fewer qubits.

A two-step downfolding strategy enhances quantum simulation efficiency.

The approach is effective for molecular systems like water in cc-pVTZ basis.

Abstract

We assess the performance of the Quantum Flow (QFlow) algorithm employing cost-effective solvers based on the unitary coupled-cluster ansatz with single and double excitations (QFlow-SD). The resulting energies are benchmarked against those obtained with an analogous QFlow formulation defined in the same active spaces but augmented by higher-rank excitations, including triples and quadruples (QFlow-SDTQ). Across all molecular systems considered, QFlow-SD exhibits close agreement with results from the canonical unitary coupled cluster with singles and doubles framework, while requiring substantially fewer qubits than the latter. For the water molecule in the cc-pVTZ basis, we further demonstrate the performance of a composite two-step downfolding strategy. In this approach, an initial coupled-cluster downfolding based on the double unitary coupled-cluster ansatz is followed by a QFlow treatment within the resulting target space, illustrating the effectiveness of combining classical downfolding with quantum flow optimization.