Density matrix search using direct inversion in the iterative subspace as a linear scaling alternative to diagonalization in electronic structure calculations

摘要: For electronic structure calculations on large systems, solving the self-consistent-field (SCF) equations is one of key bottlenecks. Density matrix search methods provide an efficient linear scaling approach for circumventing traditional O(N3) diagonalization procedure SCF equations. The conjugate gradient density (CG-DMS) method a successful implementation this approach. An alternative method, QN–DMS, employs direct inversion in iterative subspace using quasi-Newton (QN) step to estimate error vector. polyglycine chains 10–100 residues, present scales linearly and 30% faster than CG-DMS. clusters up 300 water molecules, shows smoother convergence, displays nearly scaling.