v3.0.5
What's Changed
Features
- Optimized default setting of nbands by @dyzheng in deepmodeling#1582
- Optimize OpenMP support to Force_Stress part in LCAO codes by @Alcanderian in deepmodeling#1559, deepmodeling#1580, deepmodeling#1589, deepmodeling#1594, deepmodeling#1596, deepmodeling#1628, deepmodeling#1635, deepmodeling#1666
- Support GPU Davidson method by @haozhihan in deepmodeling#1566
- Add CUDA/multi-device support for function cal_force_nl by @denghuilu in deepmodeling#1595, deepmodeling#1641, deepmodeling#1653, deepmodeling#1661
- Switch to old convergence criterion for new relaxation scheme by @wenfei-li in deepmodeling#1626
- Add parameter to control HSR printing interval during md by @wenfei-li in deepmodeling#1630
- Support read number of element types from the STRU file. by @sunliang98 in deepmodeling#1640
- Support K point parallel of Davidson method by @haozhihan in deepmodeling#1645
- EXX numerical type is distinguished by new input parameter "exx_real_number" by @PeizeLin in deepmodeling#1665
- Add i/o & mixing of tau for mGGA by @wenfei-li in deepmodeling#1682
Build and Compilation
- Modify codes for compiling with Intel icx by @caic99 in deepmodeling#1575
- Fix Makefile by @Qianruipku in deepmodeling#1634
- Fix: add missed headers for compiling with gcc-12 by @caic99 in deepmodeling#1636
- Build(conda): add DeePKS and Libxc support. by @caic99 in deepmodeling#1637
Refactor and Code-cleanup
- Move KEDF of OFDFT to
module_hamilt/of_pw
by @sunliang98 in deepmodeling#1579 - Remove redundant codes in src_io by @hongriTianqi in deepmodeling#1581, deepmodeling#1624, deepmodeling#1663, deepmodeling#1664, deepmodeling#1679, and by @denghuilu in deepmodeling#1621
- Remove unnecessary device memory operations by @denghuilu in deepmodeling#1620
- Removed unused input parameter tot_magnetization by @wenfei-li in deepmodeling#1651
CI and Test coverage
- CI: add vtune into Intel docker image by @caic99 in deepmodeling#1571
- Add e2e tests for vdwd3 and minor clean up by @xinyangd in deepmodeling#1639
- Unit testing for
K_Vectors::renew()
by @ruiyiQM in deepmodeling#1646 - Add timeout to coverage test command. by @caic99 in deepmodeling#1652
- CI: update vdw3 example by @hongriTianqi in deepmodeling#1654
- CI: update 250_vdw3abc by @hongriTianqi in deepmodeling#1657
- CI: update jd of 250_NO_KP_CR_VDW3ABC by @hongriTianqi in deepmodeling#1660
- Add a unit test for read_kpoints ("Line_Cartesian") by @ruiyiQM in deepmodeling#1675
- Add UT of input by @hongriTianqi in deepmodeling#1683
Bug fix
- Fix memory leak in energy_dos by @caic99 in deepmodeling#1576
- Fix the stress calculation in tau_xc for SCAN0 by @ouqi0711 in deepmodeling#1585
- Fix exx + relaxation calculation by @xinyangd in deepmodeling#1593
- Fix: avoid memory leak in
Local_Orbital_Charge
by @caic99 in deepmodeling#1629 - Fix: stresses are output too close when they are too large by @Qianruipku in deepmodeling#1642
- Fix: symmetry = -1 will lead to unnecessary symmetry operations by @dyzheng in deepmodeling#1648
- Fix : occupation matrix control for DFT+U by @wenfei-li in deepmodeling#1656
- Fix & Perf: wrong sdft and sKG results by @Qianruipku in deepmodeling#1680
- Rearrange output flags for scf and nscf calculation by @hongriTianqi in deepmodeling#1686
Documentation
- Add ase doc by @hongriTianqi in deepmodeling#1433
- Update FAQ by @hongriTianqi in deepmodeling#1583
- Update easy_install.md by @xinyangd in deepmodeling#1586
- Update RI related documentation by @xinyangd in deepmodeling#1587
- Update contribution guidelines. by @caic99 in deepmodeling#1590
- Update the default behavior of
cal_force
andcal_stress
by @hongriTianqi in deepmodeling#1622
New Contributors
- @ruiyiQM made their first contribution in deepmodeling#1646
Full Changelog: v3.0.4...v3.0.5