Unbiasing fermionic quantum Monte Carlo with a quantum computer.

Area of Science:

• Computational Chemistry
• Quantum Computing
• Quantum Many-Body Physics

Background:

• Interacting many-electron problems are computationally intensive, hindering accurate predictions of quantum system properties.
• Fermionic quantum Monte Carlo (QMC) methods are powerful but face biases due to computational constraints.
• Classical computation limits the flexibility of constrained QMC, impacting accuracy.

Purpose of the Study:

• To develop a hybrid quantum-classical approach to mitigate biases in constrained QMC.
• To leverage quantum computation to improve the accuracy of electronic structure calculations.
• To explore a new pathway for achieving quantum advantage in computational chemistry.

Main Methods:

• Combining constrained quantum Monte Carlo (QMC) with quantum computation.
• Experimental implementation using up to 16 qubits.
• Application to chemical systems with up to 120 orbitals.

Main Results:

• Successfully reduced biases in constrained QMC calculations.
• Achieved accuracy competitive with state-of-the-art classical methods.
• Demonstrated the largest chemistry simulations performed using quantum computers to date.
• Avoided burdensome error mitigation techniques.

Conclusions:

• The proposed hybrid quantum-classical model offers a viable alternative to variational quantum eigensolver for electronic structure problems.
• This approach provides a path towards practical quantum advantage without requiring perfect ground-state wavefunction preparation and measurement.
• The method effectively addresses the computational challenges posed by interacting many-electron systems.