CASUS Institute Seminar, Dr. Kai Trepte, postdoctoral research fellow, SUNCAT Center for Interface Science and Catalysis, Stanford University (USA)
There exist a vast variety of fast density functional theory (DFT) codes, which focus on efficiency and speed. Typically, such DFT codes are written using highly efficient pragmas which are, however, hard to read, to understand, to extend or to manipulate in order to establish complex workflows.
With NextDFT we aim for educational DFT pragmas which allow to implement DFT by anyone, i.e., students and younger researchers, within a reasonable amount of implementation time. We will outline the problems of established DFT codes and closely related problems in education. In addition, we give first insights in prototype codes following the NextDFT ideas.