Quantum algorithms that simulate physical systems involve the measurement of observables (or "expectation values") directly on a quantum computer. However, the process of calculating these expectation values from the raw result given by the quantum computer introduces a classical computational overhead which is easily underestimated. As quantum computers (and the problems they can be applied to) continue to grow in scale, we need to start thinking about how to process large amounts of output data efficiently.
Description
In algorithms such as the Variational Quantum Eigensolver (VQE), observables (e.g. the energy of a molecule) are computed by breaking them up into sums of “Pauli strings”, which are directly measurable on a quantum computer. As problem sizes grow, so do the number of Pauli strings - however, many of these strings are simultaneously measurable with a single quantum circuit, meaning that the bottleneck in calculating expectation values is often the classical post-measurement processing. We will try and come up with some clever tricks to make this process as efficient and scalable as possible.
A program which, given the measurement results from a list of quantum circuits, can calculate the expectation values of a (much longer) list of corresponding Pauli strings.
Abstract
Quantum algorithms that simulate physical systems involve the measurement of observables (or "expectation values") directly on a quantum computer. However, the process of calculating these expectation values from the raw result given by the quantum computer introduces a classical computational overhead which is easily underestimated. As quantum computers (and the problems they can be applied to) continue to grow in scale, we need to start thinking about how to process large amounts of output data efficiently.
Description
In algorithms such as the Variational Quantum Eigensolver (VQE), observables (e.g. the energy of a molecule) are computed by breaking them up into sums of “Pauli strings”, which are directly measurable on a quantum computer. As problem sizes grow, so do the number of Pauli strings - however, many of these strings are simultaneously measurable with a single quantum circuit, meaning that the bottleneck in calculating expectation values is often the classical post-measurement processing. We will try and come up with some clever tricks to make this process as efficient and scalable as possible.
Members
hvallury@student.unimelb.edu.auDeliverable
GitHub repo
TBA