Efficient materials modelling on HPC with BigDFT

BigDFT is an open source density functional theory code which uses a Daubechies wavelet basis set which facilitates optimal features of flexibility, performance and precision. In addition to the traditional cubic-scaling DFT approach, BigDFT also contains an approach which scales linearly with the number of atoms, enabling DFT calculations of large systems containing many thousands of atoms which were impractical to simulate even very recently. BigDFT consists of a package suite with a wide variety of features, from ground-state quantities to excited state quantities based on time-dependent DFT and constrained DFT, to potential energy surface exploration techniques. It uses dual space Gaussian type norm-conserving pseudopotentials including those with non-linear core corrections, which have proven to deliver all-electron precision. Its flexible Poisson solver can handle a number of different boundary conditions including free, wire, surface, and fully periodic, while it is also possible to simulate implicit solvents as well as external electric fields. Finally, BigDFT has been designed to exploit HPC from the outset, with a hybrid MPI/OpenMP approach, as well as efficient exploitation of GPUs for hybrid functional calculations.

ENCCS is based in Sweden and provides training and support for accessing and using European supercomputers for free.