Skip to content

Yet another DFT+U+GPU problem in forces

giannozz requested to merge giannozz/q-e:force_hub_again into develop

State of the art of noncolinear DFT+U[+V] on GPU after this patch:

  • Norm-conserving U, everything seems to work
  • US and PAW, stress and forces still wrong
Edited by giannozz

Merge request reports