First principles calculations using density matrix divide-and-conquer within the SIESTA methodology

Abstract

The density-matrix divide-and-conquer technique for the solution of Kohn-Sham density functional theory has been implemented within the framework of the SIESTA methodology. Implementation details are provided where the focus is on the scaling of the computation time and memory use, in both serial and parallel versions. We demonstrate the linear scaling capabilities of the technique by providing ground state calculations of moderately large insulating, semi-conducting and (near)metallic systems. [TODO - SHOW SPECIFIC EXAMPLES AND RESULTS]. This linear scaling technique has made it feasible to calculate the ground state properties of quantum systems consisting of tens of thousands of atoms with relatively modest computing resources. Extensions to the divide-and-conquer technique have also been studied. Where the smoothing of subsystem boundaries is seen to reduce the errors involved when atoms enter and leave subsystems in molecular dynamic simulations. Using mixed basis sets to increase the efficiency of the calculations and finally using the divide-and-conquer method to seed the calculation using the order-N Kim-Mauri-Galli functional minimization method.