Advances in highly accurate methods for calculating the ground and excited state energies of large assemblies of atoms are allowing ever more sophisticated studies of materials at the most fundamental level. Such calculations make it possible to predict a wide range of properties of critical importance in many branches of science and engineering, and serve as benchmarks against which other more approximate approaches may be judged. We use a wide range of electronic structure techniques, including quantum Monte Carlo methods, local density functional theory, many-body perturbation theory and, to a lesser extent, Hartree-Fock and configuration interaction methods. Our main focus is on the ground and excited state properties of solids, including the properties of point defects and surfaces. Our approach is targeted at predicting the physical properties of real systems using no input beyond that of the atomic numbers of the constituent atoms.