Exjobbspresentation: Oskar Leinonen, Kemi och biokemi
Parameter Transfer for Variational Quantum Algorithms on Near-Term Quantum Computers - An assessment of performance for potential energy surface calculations
Quantum computers show great potential for revolutionizing our ability to perform certain calculations. It has already been shown that current Noisy Intermediate-Scale Quantum (NISQ) computers can outperform conventional computers for some artificial problems. One area that would benefit strongly from quantum computations is quantum chemistry. The Variational Quantum Eigensolver (VQE) algorithm is a quantum chemistry algorithm suitable for contemporary NISQ devices. VQE optimizes a parameterized wavefunction ansatz to find the ground state energy of a molecule. In this thesis, the idea of using the optimal parameters of one molecular geometry as an initial guess for similar geometries is assessed. A significant speedup is obtained in the case of an ideal quantum computer. However, addition of noise to the simulation decreases the gain in computational speed and potentially causes the parameter transfer scheme to reduce accuracy. Furthermore, the results show that the effectiveness of parameter transfer is highly dependent on the choice of classical optimizer, which also has a huge impact on the overall performance of the VQE algorithm.
Examinator: Martin Rahm
The presentation can be viewed through Zoom https://chalmers.zoom.us/j/61094058851
Please turn off your own camera and microphone when joining by link.