- SeqQuest publicly available as computational tool on web portals: nanoHUB and memsHUB.
- Lanthanides (PBE), complete PBE library from H through Bi
- Current production version: SeqQuest-2.68 (June-2018)
- Expanded//unified SLICs geometry constraints (+bond,layers,multiple)
- Fix idle-proc solver bug
- Transferred to git-repo
- Previous production version: SeqQuest-2.67 (Aug15-Dec15)
- Label-selected atoms in specifying constraints in geometry
- Graceful NEB exit upon stall (image scf failure)
- Gamma-point add-on in post-scf
- Bugfixes (null bands, fft exception, et al.)
- Earlier production version: SeqQuest-2.66 (24Apr15)
- Primary purpose and effect: reduce disk I/O footprint in parallel runs
- Option to input geometry from separate file “lcao.geom_in” (see geomfile option in setup phase)
- Turn scf-guess default to OFF for complex (to avoid known error in gesser)
- Old version: SeqQuest-2.65d (Aug13-Dec14) Version history
- C6-based van der Waals corrections (Grimme’s D2, and Kim/Goddard ULG)
- Improved cell optimization, and more constraints for cell optimization
– more stable (and faster) cell minimizations, more means to constrain cells - Default settings recalibrated – with numerical differences
- Grid range parameter extend from 2d-7 to 2d-8
– This will lead to slight numerical differences from previous versions
– To recover pre-2.65 results, set “cutgrd” to 2.d-7
– Needed to more dependably converge stress/cell-opt and surface energy calculations. – Test suite updated to reflect changes in defaults - Other cutoffs (e.g., force convergence, scf convergence) also modified
- Grid range parameter extend from 2d-7 to 2d-8
- Bugfixes/patches:(a) 24Sep13 – NEB+LMCC fix, and output cell enthalpybugfix: lmcc crystal reference file to all image masters (not just NEB master)
diagnostic: Output enthalpy (E+PV) in cell optimizations (and when requested)(b) 5Feb14 – broken BLAS calls, band labelsbugfix: fixed broken BLAS calls in some new branches of cell opt
diagnostic: fix band labeling in band structure output(c) 9May14 – fix vdw-data inputbugfix: job-stopping bugs in reading vdw parameters from input file(d) 15May14 – fix band structure calculationbugfix: intermittent seg-fault/crazy-results from a coding error
- Old version: SeqQuest-2.64 (Jul13)
- k-distributed parallelism, including solvers, enabling larger cells with more k-points
– 500-atoms with many k-points routine, 1000-atoms practical - Built-in (post-scf) band structure facility (thx to AHE and ACP)
- k-distributed parallelism, including solvers, enabling larger cells with more k-points
- Old version: SeqQuest-2.63b (May12-Feb13)
- Parallel performance: k-parallel scf, and k-point distributed memory
– enablss large number of k-points for large cell sizes in parallel - Simple double-layer (charged slab) boundary conditions
- Bugfixes/patches:(a) 16Jan13 – bugfix: parallel solver memory distribution for small 4-proc problems(b) 26Feb13 – bugfix: harmonized stripe parallel vs. full in different parts of code
- Parallel performance: k-parallel scf, and k-point distributed memory
- Old version: SeqQuest-2.62 (Feb12)
- Improved parallelism: stripe-parallel atomic 2-, and 3-center matrix elements
- Axis-based volume constraints (vgfixX)
- Fixed-shape (i.e. isotropic) cell-volume optimization
- spin-AM05 (Rudy Magyar)
- Web pages expanded:
- Atom library updates:
- New: PBE and LDA: “long-tail” carbon atoms (for long-range, e.g. graphite, vdW calculations)
- New: Lanthanides for PBE (full library from H through Bi) [Mar2014]
- Expanded: AM05 (Check library)
- Replaced: various (what is free comes with no guarantee—check file documentation)
Coming “soon”
- Local exact exchange, HSE (coming in 2.69)
- More parallelism (memory distribution, I/O-reducing)
Wish list
- TD-DFT, linear response, GW
- Electrochemistry features – coupled DFT-solvation models
- More parallelism, all levels: Fine-grain scalable, threads
- Parallel eigensolvers (scalable for QM regime)
- Spin-orbit coupling
- Anti-ferromagnetism
- Expanded AM05 (steadily) and BLYP (not-so-much) atom library
- Projector operators and atomic hyperfine
- Linear scaling solver in production code
- QM/MM coupling
- f-orbitals
- Performance enhancements
- “Dynamic” one-center analytic/grid corrections
- New non-local integrals
- New angular integrals for local three-center integrals