From e0f2c994670f905861cac4827b22d71b74778ea9 Mon Sep 17 00:00:00 2001 From: ouqi0711 Date: Sun, 4 Jun 2023 10:57:18 +0800 Subject: [PATCH] Update faq.md (#2581) for nbands setting. Co-authored-by: Mohan Chen --- docs/community/faq.md | 3 +++ 1 file changed, 3 insertions(+) diff --git a/docs/community/faq.md b/docs/community/faq.md index 2109c5ee13..4d3c7e8848 100644 --- a/docs/community/faq.md +++ b/docs/community/faq.md @@ -56,6 +56,9 @@ Run `/usr/bin/time -v mpirun -n 4 abacus`, and locate "Maximum resident set size The tag `smearing_sigma` is used for SCF calculation, and does not affect NSCF calculation. The tag `dos_smearing` is only used for plotting density of states, which does affect SCF or NSCF results. So `smearing_sigma` should not be set in dos calculation. +**11. How to set `nbands` and `ncpus`?** + +For both pw and LCAO calculations, the default value for `nbands` can be found [here](https://abacus.deepmodeling.com/en/latest/advanced/input_files/input-main.html#nbands). Note that the number of CPUs called for a parallel job (i.e., the number specified after the command `mpirun -n`) should be smaller than `nbands`, otherwise the job will fail with an error message `nbands < ncpus`. Note also that for LCAO calculations, `nbands` should always be smaller than `nlocal`, i.e., the number of the atomic orbital basis of the system. [back to top](#frequently-asked-questions)