Dr Michele Guidolin
Michele works on improving the speed and scalability of Met Office codes on the supercomputer.
Michele is a member of the High Performance Computing Optimisation team, whose main work includes optimising the performance of the Unified Model. His aims are:
- To improve the scalability of the UM by optimising the communication algorithms used.
- To speed-up the models by making the best use of multi-threading and vectorization techniques.
- To optimise the models to run on new computer architectures.
Prior to joining the Met Office in 2015, Michele worked as a Research Fellow at the University of Exeter. Here he developed a framework for the parallelisation of Cellular Automaton (CA) algorithm used for the prediction of flooding. He has experience in developing and optimising scientific code on various HPC systems (Cray XC40, IBM power series and Linux clusters) and architectures (X85-64 CPUs, NVIDIA GPUs and Intel Xeon Phi).
Michele received his BSc and MSc in Computer Science from CA’Foscari Venice University, before obtaining his PhD in Computer Science from University College Dublin (UCD) in 2010 for his work looking at the performance of Grid computing middleware and programming models.