v3.1.1
What's Changed
Features
- Add single precision support for multi-device ops by @denghuilu in deepmodeling#1729
- Add multi-device support for HPsi(Meta) by @denghuilu in deepmodeling#1733
- enables NSCF calculation for DeePKS by @ouqi0711 in deepmodeling#1746
- enables multi-k bands training for DeePKS by @ouqi0711 in deepmodeling#1731
- Add ROCm implementation for ABACUS! by @denghuilu in deepmodeling#1726
- Add pure single precision support for ABACUS(PW Basis) by @denghuilu in deepmodeling#1763
- Add single precision support for ROCm operators by @denghuilu in deepmodeling#1767
- OpenMP for Gint_k::folding_vl_k by @Alcanderian in deepmodeling#1764
- mixing of locale in DFT+U by @wenfei-li in deepmodeling#1743
- enable relax and cell-relax for lj/dp pot by @YuLiu98 in deepmodeling#1762
- OpenMP for OperatorLCAO::folding_fixed by @Alcanderian in deepmodeling#1772
- add input parameter td_print_eij and td_edm for tddft by @lyb9812 in deepmodeling#1789
- OpenMP for Local_Orbital_Charge::cal_dk_k by @Alcanderian in deepmodeling#1773
Build and Compilation
- add an option to enable native compilation optimization by @njzjz in deepmodeling#1724
- use Release for default build type by @caic99 in deepmodeling#1722
- trim cmake flags by @caic99 in deepmodeling#1784
Refactor and Code-cleanup
- Remove unnecessary transpose operations in CG-subspace by @haozhihan in deepmodeling#1723
- Remove unnecessary transpose operations in Davidson by @haozhihan in deepmodeling#1727
- Remove all CUDA APIs from modules by @denghuilu in deepmodeling#1715
- Add device template for member functions by @denghuilu in deepmodeling#1707
- New RPA interface with LibRI by @Srlive1201 in deepmodeling#1717
- Refactor the code structure of multi-device operators by @denghuilu in deepmodeling#1749
- modify the default of scf_nmax to 100 by @pxlxingliang in deepmodeling#1765
- electric field in tddft by @lyb9812 in deepmodeling#1774
- port some libm source file to abacus: sincos, exp, cexp. and… by @Alcanderian in deepmodeling#1780
- change folder name src_io to module_io by @hongriTianqi in deepmodeling#1793
- new memory record interface by @dyzheng in deepmodeling#1794
CI and Test coverage
- Add device template for static members of class DiagoIterAssist by @denghuilu in deepmodeling#1713
- add ut for default_2 and check of input.cpp by @hongriTianqi in deepmodeling#1725
- add UT for winput.cpp by @hongriTianqi in deepmodeling#1730
- Add a unit test for klist.cpp ("set_kup_and_kdw_after_vc") by @ruiyiQM in deepmodeling#1736
- add UT for output.cpp by @hongriTianqi in deepmodeling#1739
- remedy updaterhok_pw_test & add ut for write_wfc(bit) by @hongriTianqi in deepmodeling#1747
- add UT for rwstream.cpp by @hongriTianqi in deepmodeling#1741
- add 2 warnings in input.cpp by @hongriTianqi in deepmodeling#1752
- update test 702_DP_Al for dp pot by @YuLiu98 in deepmodeling#1759
- add unit tests for module_md by @YuLiu98 in deepmodeling#1771
- UnitTest for realArrayAlloc by @SeptemberMy in deepmodeling#1788
- UT of complexArrayxAlloc by @hongriTianqi in deepmodeling#1786
- Delete ntype by @dyzheng in deepmodeling#1770
- add e2e tests for get_S out_proj_band out_pot and out_dm1 by @hongriTianqi in deepmodeling#1732
- add e2e example of W90 in pw by @hongriTianqi in deepmodeling#1734
- add an example of spin4 & out_mat_hs by @hongriTianqi in deepmodeling#1748
- fix 207_NO_KP_OHS_SPIN4 by @hongriTianqi in deepmodeling#1757
- UnitTest for ZEROS by @SeptemberMy in deepmodeling#1798
Bug fix
- Fix memory leak within pseudopot_cell_vnl::getvnl function by @denghuilu in deepmodeling#1714
- Fix RI related compilations errors when using clang compiler by @xinyangd in deepmodeling#1721
- fix memory leak in
Local_Orbital_Chargeby @caic99 in deepmodeling#1720 - add atomic magnetism output in mulliken.txt by @dyzheng in deepmodeling#1738
- correct the INPUT in examples/dft_plus_u/NiO by @pxlxingliang in deepmodeling#1745
- fix para nelec in gatefield by @YuLiu98 in deepmodeling#1751
- fix bugs of multi-kpoints for tddft and add autotest by @lyb9812 in deepmodeling#1754
- mulliken charge for spin-down with gamma_only mode by @dyzheng in deepmodeling#1755
- bug of extreme large number of k-points during skipping non-occupied bands by @dyzheng in deepmodeling#1769
- two too-long timer function print info by @Liu-RX in deepmodeling#1768
- use
static_castfor init list by @caic99 in deepmodeling#1783 - initial magnetization with both element and atoms by @dyzheng in deepmodeling#1778
- update version-number output to v3.1 by @dyzheng in deepmodeling#1792
Documentation
- fix a typo in quick_start/input.md by @hongriTianqi in deepmodeling#1719
- add guidelines on runtime errors and debugging. by @caic99 in deepmodeling#1737
- add a symlink to citation conforming with GitHub by @caic99 in deepmodeling#1750
- add instrutions for launching abacus from a custom dir. by @caic99 in deepmodeling#1756
New Contributors
- @njzjz made their first contribution in deepmodeling#1724
- @SeptemberMy made their first contribution in deepmodeling#1788
Full Changelog: v3.1.0...v3.1.1