Abstract
We present a method for electronic structure calculations that retains all of the advantages of real space, but addresses the major weakness of a regular grid, i.e., its inability to treat some regions of space with more resolution than others. The computations are carried out on a regular mesh in curvilinear space, which allows natural and efficient decomposition on parallel computers, and effective use of iterative numerical methods. A novel feature is the use of error analysis to optimize the curvilinear grid for highly inhomogeneous electronic distributions. We report accurate all-electron calculations for H 2, O, and O 2.
Highlights
We present a method for electronic structure calculations that retains all of the advantages of real space and addresses the inherent inefficiency of a regular grid, which has equal precision everywhere
To achieve good load balance, computational complexity and memory requirements must be evenly divided among processors, a task made very difficult by complex bases like linearized augmented plane waves (LAPW) and LMTO
A regular grid in real space suffers from the same drawbacks as a plane wave basis, i.e., it has the same resolution in every region of space
Summary
Real-space electronic structure calculations on parallel computers We present a method for electronic structure calculations that retains all of the advantages of real space and addresses the inherent inefficiency of a regular grid, which has equal precision everywhere. The computations are carried out on a regular mesh in curvilinear space, which allows natural and efficient decomposition on parallel computers, and effective use of iterative numerical methods.
Sign-in/Register to view highlights & summary 
Accepted Version (
Free)
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call