The remaining time is spent pursuing computational science research. In 2018, for example, I investigated the benefits of using one-sided communications within the Met Office NERC Cloud model. The findings of this work were presented at the CUG 2018
conference. I have also contributed to the ExaFLOW
project, which concerned the need to prepare CFD codes for the move to exascale platforms. This work involved the in-depth profiling of the Nektar++
code using a varierty of profiling tools such as CrayPat
and Arm MAP
. In addition, I developed a suite of Python scripts to post-process CrayPat diagnostic data in order to produce communications mosaics for runs involving 1000+ MPI processes and to generate flame graphs
that analysed IO performance.