Phase estimation with partially randomized time evolution
read the original abstract
Quantum phase estimation combined with Hamiltonian simulation is the most promising algorithmic framework to computing ground state energies on quantum computers. Its main computational overhead derives from the Hamiltonian simulation subroutine. In this paper we use randomization to speed up product formulas, one of the standard approaches to Hamiltonian simulation. We propose new partially randomized Hamiltonian simulation methods in which some terms are kept deterministically and others are randomly sampled. We perform a detailed resource estimate for single-ancilla phase estimation using partially randomized product formulas for benchmark systems in quantum chemistry and obtain orders-of-magnitude improvements compared to other simulations based on product formulas. When applied to the hydrogen chain, we have numerical evidence that our methods exhibit asymptotic scaling with the system size that is competitive with the best known qubitization approaches.
This paper has not been read by Pith yet.
Forward citations
Cited by 7 Pith papers
-
Quantum simulation of nanographenes and Trotter error cancellation
Trotter error cancellation for energy gaps in nanographene simulations yields an order-of-magnitude reduction in quantum circuit depth, enabling chemical-accuracy estimates with fewer than 3.2×10^7 Toffoli gates for l...
-
Sample-Based Quantum Diagonalization with Amplitude Amplification
SQD-AA reduces total query complexity by more than 100x on model distributions and achieves the lowest T-gate counts with 3-4 orders shallower circuits than iQPE for molecular examples.
-
Quantum simulation of nanographenes and Trotter error cancellation
Trotter error cancellation in nanographene simulations reduces circuit depth by about 10x for quantum phase estimation of energy gaps to chemical accuracy in the Pariser-Parr-Pople model.
-
When is randomization advantageous in quantum simulation?
Randomized sparse-QSVT reduces gate counts by up to 10x for inhomogeneous many-term Hamiltonians at moderate error (around 10^{-3}), but deterministic QSVT becomes cheaper for higher precision.
-
A unified quantum computing quantum Monte Carlo framework through structured state preparation
Structured state preparation in QCQMC improves energy accuracy over pure variational methods across molecular, condensed-matter, nuclear, and graph problems.
-
How to use quantum computers for biomolecular free energies
A two-fold quantum embedding strategy combined with machine learning integrates accurate quantum-mechanical energies into free energy calculations for biomolecular complexes and analyzes requirements for quantum compu...
-
Space and Time Cost of Continuous Rotations in Surface Codes
Catalyst towers reduce runtime and spacetime volume for continuous rotations in surface codes at small and medium distances in phase oracle and variational state preparation circuits for option pricing.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.