Huanjing Gong

I have tested three orbitals generated by various reference strategies. Only the last one shows numerical stabilty. all occupied + 4 unoccupied bands  SZ occupied bands, DZP occupied +...

@kirk0830 All PBE have converged. I will relay your suggestion next week.

I have attempted to find out two possible causes. It was likely numerical instability in local RI. Tests are attached here. [basisfix.tar.gz](https://github.com/user-attachments/files/16594985/basisfix.tar.gz) `issue.300714.out` is the original issue, in which `ETOT`...

Thanks, I get it. While HF calculation doesn't work by "occ+4", it perfoms better in PBE. Previously, I have encountered band structure of two element that is not consistent with...

The same error with OMP_NUM_THREADS=1. 

> @AroundPeking Has this issue been solved? If not, what is suggested for improving the performance of ABACUS later? Thank you. @hongriTianqi I have not solved it yet. In fact,...