Michele Guidolin Email & Phone Number

Exeter, England, United Kingdom

Exeter, United Kingdom

In my role in the HPC optimisation team, I examined and improved the performance and scalability of the Met Office's main codes on the current and new supercomputers.I have been part of the team creating the operational requirements for the HPC2020+ procurement of the new Met Office supercomputer. I selected the benchmark models that were used to evaluate its performance. I was also involved in the evaluation of the proposed solutions and dialogue with the tenderers. After the procurement, I helped teams of SSEs and scientists in porting our models onto the new supercomputer and I was the “technical assurance” role in the HPC porting project board.

Designed a framework for the parallisation of Cellular Automaton (CA) algorithm used for the prediction of flooding.Createde a new rapid flood model that uses CA technique for the CADDIES project. Worked of optimisation of tidal farm layout and on pipe networks using Genetic Algorithms.

Designed and developed CWSNet.Investigated the characteristics and performance of various iterative methods and direct methods for solving linear systems of equations generated by hydraulic solvers. I Developed a conjugate gradient method (CGM) that performs the computation on a GPU using the CUDA toolkit.

Phd, Computer Science

Activities and Societies: President International Student Society (ISS); Treasure International Student Society; Laboratory Teaching.

Msc, Computer Science

Developed and tested an application (GAMMS) for computing gravitational fields that implemented a multigrid algorithm with adaptive mesh.

Bsc, Computer Science

Studied different kinds of HPC architectures and developed and tested a scheduler in C for a distributed compiler. The title of the thesis.