Hi all,
Tomorrow Jhonathan will talk about hybrid quantum-classical variational
algorithms for chemistry. See below for the title and abstract. Also, if
you have a conflict with one of the possible new group meeting times and
haven't replied, please do!
See you there,
Ian
-----------------
Title: Perspectives on hybrid quantum-classical variational algorithms for
chemistry
Abstract: Quantum computers have arrived and are here to stay. The
technological race between the experimental groups in companies and
universities will soon deliver the first medium size non-error corrected
quantum computers. These machines are projected to be able to run a few
thousands gates in about 50 qubits, just about the size of the biggest
molecular systems that can be acurretely studied using full configuration
interaction (FCI) or density matrix renormalization group (DMRG) approches.
Among all the quantum algorithms for quantum chemistry, hybrid
quantum-classical methods, such as the variational quantum eigensolver
(VQE), are in a better position to take advantage of the new quantum
devices due to their lower requirement in coherence time and relative
robustness to noise.
In this talk, I will briefly review the VQE algorithm and its
implementation using a unitary coupled cluster (UCC) ansatz. I will discuss
the method from the perspective of the static vs dynamic correlation
paradigm, commonly used to guide the design of multireference approaches in
quantum chemistry. I will describe how the same concepts can be used to
generate new hybrid quantum-classical schemes that can be implemented on
medium-sized quantum computers. I will illustrate the methodology with two
examples: 1) generalization of UCC to entangled reference states and 2) UCC
with orbital optimization.
Show replies by date