anout lanczos

[hakuna]

Dear all:
compiling the TST code here into VASP, all are oK except running the laczos.
the problem is "FORTRAN STOP", it seems the compiler, PGI, goes wrong, but the OUTCAR said "IOPT > 0 to use the laczos method" , thus we put a "IOPT=2" in INCAR, vasp runs happly.....
Why "IPOT>0" can steer the laczos?
In addition, the IFORT compiled vasp does not show such problem.

[graeme]

The issue here is that you must use one of our geometry optimizers to run the dimer and lanczos routines. Also, make sure to set IBRION=3 and POTIM=0 so that our optimizers do not conflict with those from vasp.