My Community

Hello VASP community,

I am currently working on a project that involves using the RMM-DIIS algorithm for molecular dynamics (IBRION=5). I have been trying to optimize my calculations and was considering the use of the NPAR, NCORE, and KPAR options to speed up the process.

However, I encountered an issue where the internal routines requested a change of the k-point set, and the error message suggested removing the NPAR tag from the INCAR file. This led me to wonder why these options might not be available or effective with IBRION=5.

Could anyone provide some insight into this? Specifically, I am interested in understanding:

Why might NPAR, NCORE, and KPAR not be available with IBRION=5?
Are there any specific parallelization requirements or limitations of the RMM-DIIS algorithm that could affect the use of these options?
Are there alternative ways to optimize the speed of calculations when using IBRION=5?
Any help or guidance would be greatly appreciated. Thank you in advance for your time and assistance.

"VASP internal routines have requested a change of the k-point set. |
| Unfortunately, this is only possible if NPAR=number of nodes. |
| Please remove the tag NPAR from the INCAR file and restart the |
| calculation."

Hi,

Thanks for posting on the VASP forum. I think you are confusing the RMM-DIIS algorithm used in the electronic minimization (ALGO=Fast or ALGO=VeryFast) with the computation of phonons (IBRION=5). Molecular dynamics calculations are activated by setting IBRION=0.

For phonon calculations (IBRION=5 or IBRION=6) with symmetry enabled (ISYM>0), the use of NCORE>1 is not supported.
For more information on the different levels of parallelization available in VASP, please consult the corresponding wiki page.