Conjugate gradient minimization of the energy functional: A new method for electronic structure calculation

Abstract
A new robust iterative method for electronic structure calculations based on a convenient adaptation of the conjugate gradient minimization of the energy functional is presented. The method is compared with some other techniques using the direct minimization of the density functional and with a more traditional Davidson approach. Numerical results for silicon and carbon are used to compare the different schemes. The new method is shown to be rapidly convergent, irrespective of the system under consideration, thus providing an efficient solution to a variety of large-scale electronic structure calculations.