RSEConUK 2019

High Performance Computing is a bit of a paradox. On one side you have leading-edge technology, cutting frontiers in using huge quantities of data... and on the other a workhorse workflow powered by FORTRAN with very few people around to progress the code. Somewhere in the middle is you, being asked to tackle projects from how to take clearer images of a black hole to making sure that the Physics Department’s core research doesn’t grind to a halt. How do you navigate between the known, solid projects and the unknown? This talk will take you through the division lines between known and unknown in HPC projects, discuss the characteristics which define each, and go through a handy set of tips on how to assess projects and see if they are a match to your (or your team’s) skill sets. Along the way share what we have learned in commercial HPC, on the open source projects we run, and drop some serious public cloud knowledge.