My research has all been concerned with the theory of condensed matter, with a strong emphasis on computer simulation. This really started at Harwell, where I learned how to do molecular dynamics simulation, and worked on fast ion conduction in uranium dioxide and other materials. In 1997, inspired by a now famous paper by Car and Parrinello, I started to put a major effort into computer modelling based on electronic structure techniques. My collaborators and students at UCL use a range of simulation techniques to tackle a number of scientific problems, including the properties of materials in the Earth’s deep interior, semiconductor nanostructures, and the surface chemistry of oxide materials. Our group belongs to several UK consortia of national supercomputer users, including the UK Car-Parrinello consortium, the Materials Chemistry consortium, and the Minerals Physics consortium.