Overview of New Features
Note: This major version update breaks compatibility with code written for QuOp_MPI 0.0.x.
Version 1.0.0 introduces an object-oriented approach to quantum variational design using the 'Ansatz' and 'Unitary' classes. Also introduced is the optional parallel computation of the objective function gradient via finite differences if using an optimisation algorithm that utilises gradient information.
These new features are detailed in the updated documentation and examples are given in QuOp_MPI/examples. These cover use of the pre-defined Quantum Approximate Optimisation Algorithm and the Quantum Walk-based Optimisation Algorithm. As well as the implementation of the Extended Quantum Approximate Optimisation Algorithm and the Quantum Alternating Operator Ansatz via direct use of the 'Ansatz' and 'Unitary' classes.
Supporting Materials
This release coincides with the submission of the article "QuOp_MPI: a framework for parallel simulation of quantum variational algorithms". To support the benchmarking results discussed in the article, the results of QuOp_MPI benchmarks on a Cray supercomputer are included in QuOp_MPI/benchmarks.