Optimised Baranyai partitioning of the second quantised Hamiltonian

TL;DR

This paper introduces an optimized Baranyai grouping method for partitioning second quantised Hamiltonians in quantum chemistry, enabling efficient measurement and diagonalisation with improved scalability and variance reduction techniques.

Contribution

The authors develop an optimized Baranyai partitioning approach for second quantised Hamiltonians, improving measurement efficiency and handling sparsity in quantum simulations.

Findings

Efficient construction of diagonalisation circuits in O(N) gates.

Handles sparsity and produces a small number of groups for linear Hamiltonians.

Spin-symmetry optimization reduces groups by a factor of 8.

Abstract

Simultaneous measurement of multiple Pauli strings (tensor products of Pauli matrices) is the basis for efficient measurement of observables on quantum computers by partitioning the observable into commuting sets of Pauli strings. We present the implementation and optimisation of the Baranyai grouping method for second quantised Hamiltonian partitioning in molecules up to CH$_4$ (cc-pVDZ, 68 qubits) and efficient construction of the diagonalisation circuit in $O(N)$ quantum gates, compared to $O(N^2)$, where $N$ is the number of qubits. We show that this method naturally handles sparsity in the Hamiltonian and produces a $O(1)$ number of groups for linearly scaling Hamiltonians, such as those formed by molecules in a line; rising to $O(N^3)$ for fully connected two-body Hamiltonians. While this is more measurements than some other schemes it allows for the flexibility to move Pauli strings and optimise the variance. We also present an explicit optimisation for spin-symmetry which reduces the number of groups by a factor of $8$, without extra computational effort.