Peng Xingliang

Large force difference when set different symemtry_prec

1

### Describe the bug When I use default symmetry_prec (that is 1e-6) and symmetry_prec 1e-5, the forces are different: symmetry_prec=1e-6: ``` ------------------------------------------------------------------------------------------ TOTAL-FORCE (eV/Angstrom) ------------------------------------------------------------------------------------------ Mo1 0.0000000000 0.0000000000 0.0000000000 Mo2...

### Background For the input coordinate, the direction of orbital may not exactly on standard direction, such as dxy may not exactly on xy platform. And when I plot the...

Unstable LCAO calculation of 004_Li128C75H100O75

15

### Describe the bug The LCAO calculation of daily test 004_Li128C75H100O75 is unstable. [004_Li128C75H100O75.zip](https://github.com/deepmodeling/abacus-develop/files/12679264/004_Li128C75H100O75.zip) At current version (20230921 develop branch), the intel compiled calculation is stable, but GNU compiled is...

The scf converge steps of dav_subspace are larger than dav

1

### Details I have tested the dav_subspace method, and compared the accuracy and performance to the dav method: - The difference of energy, force, virial between two methods is acceptable....

### Describe the bug I use dav_subspace to do SCF, and ABACUS throw error: ``` abacus_pw: /public/home/abacus/abacus-dy/source/module_hsolver/kernels/dngvd_op.cpp:77: void hsolver::dngvd_op::operator()(const base_device::DEVICE_CPU *, const int, const int, const T *, const T...

intel CI test

3

### Describe the Testing Issue In this issue, I will update the integrate and unit tests in intel. ### Additional Context _No response_ ### Task list for Issue attackers (only...

The voltage of NaCrO system is not correct

2

### Describe the bug The order of voltage between system NaCrO2/Na0.5CrO2 and Na0.5CrO2/CrO2 is different from that calculated by VASP and QE. system | V_ABACUS(eV) | V_VASP (eV) | V_QE(eV)...

### Describe the bug With the latest commit `c61a66eaabf0` (8-24), the dcu ABACUS is compiled failed with error: ``` clang-15: warning: argument unused during compilation: '-amdgpu-function-calls=false' [-Wunused-command-line-argument] [ 90%] Linking...

The magnetic moment is wrong when set nspin=4 but noncolin=0

1

### Describe the bug If one sets the initial x y z magnetic moment in STRU, and set nspin=4 in INPUT but not set noncolin, ABACUS can not deal with...

### Describe the bug For example, if one system has 80 k points, and when setting kpar=4, will the band information of only 20 k points are outputted in running_xxx.log...