Paradigm shift of materials design by computer simulation - From explanation to prediction

Starting from alchemy, in the history of materials development, theory always has had a role to explain experimental findings. This situation is completely different from the field of elementary particle physics, where theoreticians predict new particles or phenomena and experimentalists follow to confirm the theoretical prediction. Since materials are complex many-body systems composed of a number of electrons and nuclei, it has been difficult to compute atomic structures and physical and chemical properties without experimental help. Recent progress, in theory and a tremendous increase in computer power, has fundamentally changed this situation. For any kind of material we know the governing equation and interaction, i.e. Schroedinger equation and Coulomb force. Therefore, we basically know everything and we can in principle solve the governing equation to predict new materials before experiment. This approach is generally called the "ab initio method" and recently has been used widely in chemistry, solid state physics, and materials science. However, it is still not sufficient enough to change the paradigm of materials design mainly because the electron exchange-correlation functional in the density functional theory is not known. In this commentary, it will be shown how to overcome this long historical condition, realize a breakthrough in theoretical materials design, and shift the paradigm from explanation to prediction.