PNNL

The Science

Solving chemistry problems with quantum systems requires developing and implementing new approaches to computations. The Quantum Flow (QFlow) algorithm provides a framework to efficiently describe correlated many-body systems on hybrid quantum-classical architectures. This implementation of QFlow highlights how the algorithm breaks down larger quantum problems into subproblems that can be solved with a modest 12-qubit equivalent system. Compared with a more computationally intensive reference approach, QFlow recovered over 95% of the calculated correlation energy. This broadens the use of lower-qubit systems for larger chemical simulations.

The Impact

While quantum computing has the potential to powerfully simulate chemistry, realistic systems are currently too complex for near-term quantum computers. QFlow enables researchers to scale up many-body quantum computing calculations to fit within the practical limits of current and anticipated systems. By demonstrating an implementation of QFlow, this work demonstrates the algorithm’s utility for studying larger-scale chemical systems and obtaining meaningful results.

Summary

Quantum computers promise powerful molecular simulations, but realistic chemistry often needs significantly more reliable qubits than near-term and early fault-tolerant devices can provide. The QFlow algorithm addresses this by breaking large correlated electronic structure problems into many small active-space subproblems. Classical high-performance computing builds effective Hamiltonians that couple the active spaces, while modest quantum resources—in this case the equivalent of 12 qubits—solve each subproblem. QFlow can use various types of quantum solvers for each subproblem. QFlow thus “harvests” quantum information from many small calculations and recombines it to approximate the full system. 

Researchers explored propane as an exemplar system for QFlow calculation. Direct treatment of propane with the cc-pVDZ basis would be a 164-qubit problem not practically solvable with current and near-term quantum systems. QFlow instead ran repeated cycles of 56,860 12-qubit active spaces, optimized 1.17 million unique wavefunction parameters, and recovered approximately 95% of a reference calculated correlation energy by the third cycle. This work demonstrates reliability for larger basis-set calculations and introduces ensemble-based averaging that enables QFlow’s application to strongly correlated systems.

Contact

Nicholas Bauman, Pacific Northwest National Laboratory, nicholas.bauman@pnnl.gov 

Karol Kowalski, Pacific Northwest National Laboratory, karol.kowalski@pnnl.gov 

Funding

This material is based upon work supported by the Embedding QC into Many-body Frameworks for Strongly Correlated Molecular and Materials Systems project, which is funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences program, the Division of Chemical Sciences, Geosciences, and Biosciences (under FWP 72689). N.P.B. and A.P. acknowledge the support from the Quantum Science Center, a National Quantum Information Science Research Center of the U.S. Department of Energy (under FWP 76213), and from the Quantum Algorithms and Architecture for Domain Science Initiative under the Laboratory Directed Research and Development program at Pacific Northwest National Laboratory (PNNL). PNNL is operated by Battelle for the U.S. Department of Energy under Contract DE-AC05-76RL01830.