Quantum Espresso v5.1.1

Incompatible changes in 5.1.1 version:

  • Regularization of DFT names for nonlocal functionals: "vdw-df3" replaced by "vdw-df-obk8" "vdw-df4" and "optb86b-vdw" replaced by "vdw-df-ob86" "rev-vdw-df2" replaced by "vdw-df2-b86r" Explanation: We have tried to unify the naming convention of the vdW-DF family of functionals built on "Dion et al., PRL 92, 246401 (2004)." Within this family, numbering should be used for different non-local pieces of the functionals, such as vdW-DF2. If functionals only change the (semi)local contribution, the choice of (semi)local functional should be attached at the end of the name, such as vdW-DF-C09 or vdW-DF-cx.

New in 5.1.1 version:

  • CP: Hybrid functionals with Wannier functions
  • Added possibility to provide structure via space-group number and Wyckoff positions (experimental)
  • Added vdW-DF-cx, Berland and Hyldgaard, PRB 89, 035412 (2014) (courtesy of Timo Thonhauser)
  • PW: TB09 meta-GGA functional (requires libxc)
  • PW: the code can capture "signals". typically sent by batch queues when allowed time is close to expire, and terminate gracefully. Experimental, to be enabled at compile time (see the user guide)
  • TDDFPT: Implemented a restart option in turbo_Davidson

Fixed in 5.1.1 version:

  • PWscf: energy related to external forces wasn't correct
  • CP: Wannier functions dynamics with nspin=2 wasn't working
  • PWscf: restart in variable-cell case wasn't working in v.5.1
  • pot_extrapolation='none' was crashing with a bogus error message
  • PHonon: the recently introduced fix to the G=0 component of the deformation potential at q=0 was not working properly
  • Kernel table for rVV10 nonlocal functional was not written to file, breaking restart and postprocessing (no problem for scf)
  • Strange compilation bug on some IBM machines
  • Minor bug plus OpenMP bug in ESM implementation
  • Small bug leading to crashes when restarting some type of constrained magnetization calculations (courtesy of Florian Thoele, ETHZ)