Shift-and-Invert Parallel Spectral Transformation Eigensolver: Massively Parallel Performance for Density-Functional Based Tight-Binding

Summary

Develops the SIPs shift-and-invert spectrum-slicing eigensolver for massively parallel architectures and demonstrates it on DFTB Hamiltonians of up to 128,000-atom carbon nanotubes, obtaining 330,000 eigenpairs in 190 s on 266,144 cores.

Keywords

Cite

@article{2016_keceli_sips_eigensolver_dftb,
  title = {Shift-and-Invert Parallel Spectral Transformation Eigensolver: Massively Parallel Performance for Density-Functional Based Tight-Binding},
  author = {Murat Keçeli and Hong Zhang and Peter Zapol and David A. Dixon and Albert F. Wagner},
  journal = {J. Comput. Chem. 37, 448–459 (2016)},
  year = {2016},
  doi = {10.1002/jcc.24254}
}