CCL: ORCA 4.0 has just been released
- From: "Frank Neese"
<Frank.Neese-x-cec.mpg.de>
- Subject: CCL: ORCA 4.0 has just been released
- Date: Tue, 14 Mar 2017 09:44:23 -0400
Sent to CCL by: "Frank Neese" [Frank.Neese###cec.mpg.de]
Dear CCLers
On behalf of the ORCA development team, I have the pleasure to inform you that
today we have released
ORCA 4.0. You can download it from the Website of the Max Planck Institute for
Chemical Energy
Conversion under: https://orcaforum.cec.mpg.de following the usual procedure.
ORCA 4.0 is a major improvement over the last public release (ORCA 3.0.3) that
happened in 2014. In the
meantime, we have worked on ORCA very extensively. The result is a long list of
new methods and
features that I will summarize below. However, a lot of work has been done under
the hood. The visible
outcome of these improvements is: a) improved performance and b) a slightly
different naming
convention for basis sets. The latter was necessary in order to streamline the
handling of basis sets in
ORCA, which, admittedly, has been somewhat confusing in the past.
ORCA has come a very long way in the past years and according to some
statistical research by now is
among the top five most used quantum chemistry packages worldwide. Our user
community has grown to
more than 12000 registered individual users but the real number must be higher,
due to ORCAs presence
in many research groups and supercomputer centers world-wide.
By now, there has been so much interest from major industry in ORCA, that we
finally founded a company
that will distribute ORCA to commercial users. The company will be operative in
the next few weeks and
will make a separate announcement with instructions for interested parties. Let
me emphasize again:
ORCA is free for academic users and will remain so. The company will give access
to commercial users
which, so far, had no access to ORCA whatsoever. There are many legal aspects
touched by such an
operation and this necessitated that we create a new end-users-license agreement
(EULA) that has to be
signed by the academic users.
A big word of thanks goes to our collaborators around the world which have
greatly helped making ORCA
better! We have thoroughly enjoyed to collaborate with these wonderful
scientists and we greatly
appreciate the insights we have gotten from them. We specifically mention Ed
Valeev, Stefan Grimme and
his team, Marcel Nooijen as well as Jiri Pittner and Ondrej Demel among many
others.
The incorporation of COSMO into the academic version of ORCA is presently
re-negotiated. For the time
being, COSMO is no longer a part of ORCA and the interface to otool_cosmo has
been disabled in ORCA
4.0. Equivalent functionality is offered by CPCM model in ORCA and typically the
results are very similar.
We have put our hearts into creating ORCA 4.0 and we very much hope that the
ORCA community will
receive it as well as earlier versions and will keep growing. We are definitely
committed to giving you the
best possible software to solve your chemical problems!
We are not asking any money or donation or anything in return for giving you
ORCA. If you like it and use
it and write scientific papers that report the results of such calculations, the
only thing we ask you is to:
P-L-E-A-S-E
Cite our original research papers and NOT just the overview article that
describes ORCA in broad terms.
Your citations give us the necessary academic reputation that we need in order
to be able to secure the
resources that are required to build ORCA. So, please take this seriously. It is
a small token of
appreciation that we ask, not more, not less.
With that I very much hope that you will enjoy using ORCA 4.0 as much as we have
enjoyed creating it! As
it is inevitable that (hopefully only small) bugs will surface, we plan another
release before the end of the
year that will contain all the things that did not make it in this release and
the bugfixes that have
happened until then.
Enjoy the wonderful art and science of quantum chemistry!
Frank Neese on behalf of the ORCA development team
New Methods in ORCA 4.0
- Linear scaling DLPNO-MP2 (RHF and UHF)
- Linear scaling DLPNO-MP2-F12 (RHF)
- Linear scaling DLPNO-CCSD(T) (the 2013 implementation is still available)
- Linear scaling DLPNO-CCSD(T) local energy decomposition scheme
- Linear scaling DLPNO-CCSD closed shell density
- Linear scaling DLPNO-CCSD(T) open shell. New restricted open-shell formulation
- Linear scaling cluster in molecule (CIM): MP2, CCSD(T), DLPNO-CCSD(T)
- Linear scaling LNO-CIM-CCSD similar to Kallay
- Linear scaling DLPNO-NEVPT2
- Linear-scaling DLPNO-NEVPT-F12
- Non-linear scaling LPNO-CCSD-F12 (DLPNO-CCSD-F12 pending)
- Non-linear scaling Mukherjee Mk-LPNO-MRCCSD(T)
- Powerful iterative configuration expansion (ICE-CI) approximation to Full-CI
- ICE-CI for large active space CASSCF calculations
- A partial PNO-EOM-CCSD method for excited states
- A partial PNO-STEOM-CCSD method for excited states
- Fully internally contracted MRCI (FIC-MRCI)
- Full TD-DFT energies and gradient for hybrid functionals
- Super-fast approximate TD-DFT: sTDA/sTDDFT of Grimme and co-workers
- PBEh-3c method of Grimme and co-workers
SCF, Gradient, Hessian
- Large performance improvements (up to factor of four) for calculations with
four center integrals
(energy and gradient)
- Improved performance with RI-J with conventionally stored integrals
- Gradient for range separated hybrids
- Gradient for range double hybrid functionals with meta GGAs
- Gradient for range double hybrid functionals with range separated functionals
- Gradient for RI-JK
- Frequencies for range separated functionals
- Stability analysis and automatic search for broken symmetry states
- Local spin analysis
- PBEh-3c method
- Fractional occupation number analysis (FOD) for detection of MR character
MDCI coupled cluster module
- All improvements for DLPNO methods as listed under New Methods
- Closed shell EOM-CCSD energies
- Closed shell STEOM-CCSD energies
- Automatic closed shell STEOM-CCSD active space selection
- EOM-CCSD(2) and STEOM-CCSD(2) approximations
- EOM-CCSD transition moments
- EOM-IP for ionized states
- EOM/STEOM-CCSD core level excited states
- ADC(2) and CC(2) methods (initial implementation)
- IP-EOM-CCSD and EA-EOM-CCSD
- COSX for EOM-CCSD and STEOM-CCSD
- Improved automatic frozen core handling
- Core-correlation in automatic basis set extrapolation
New automatic code generated AUTOCI module
- RHF CISD
- RHF CCSD
- UHF CISD
- UHF CCSD
- ROHF CISD
- ROHF CCSD
- FIC-MRCI and FIC-CEPA/0
CASSCF, NEVPT2 and MRCI
- Detailed tutorial showing CASSCF/NEVPT2 usage
- Improved convergence in CASSCF
- Partially contracted NEVPT2
- Linear scaling NEVPT2
- Automatic implementation of ab initio ligand-field theory in CASSCF
- ICE-CI for large active space CASSCF calculations
- Active space constraints and external orbital manipulations in CASSCF
- MREOM-CCSD (also with SOC)
- Local spin analysis for CASSCF
- Accelerated CI (ACCCI) for more efficient CI step in CASSCF
- Fragment decomposition of the spin-spin interaction
- Cumulant approximation for NEVPT2
- New and improved DMRG interface
- ACCCI as CIStep for FIC and DLPNO-NEVPT2
- Determinant based analysis of CASCI states and printing of the same
- Explicitly correlated RI-FIC-NEVPT2 (NEVPT2-F12)
TD-DFT and ROCIS
- Full TD-DFT for hybrid functionals
- Gradient for full TD-DFT with hybrid functionals
- TD-DFT/TDA gradient with range separated functionals
- ROCIS magnetic properties (hyperfine, g-tensor, ZFS tensor, MCD)
- ROCIS-RIXS spectra
- PNO-ROCIS for spectacular performance improvements
- Super-fast approximate TD-DFT: sTDA/sTDDFT
- Natural transition orbitals in TD-DFT and ROCIS
Miscellaneous
- GIAO implementation for NMR chemical shifts. Various aproximations (RIJOCOSX,
RIJK, )
- New Handling of basis set names. Now fully consistent with TurboMole
def2-defaults (including
ECPs) SARC basis sets separately available
- New reading of basis sets and ECPs together.
- New input handling for ANO basis sets
- New correlation consistent basis sets added
- New SARC basis sets for the lanthanides; good for correlated calculations
- New ANO-RCC basis sets added
- Improved frozen core handling in correlation calculations.
- Improved automatic auxiliary basis set generation
- Corrections for low-frequency modes in thermochemistry
- New and improved NBO interface
- CPCM and improved SMD solvent models
- Intrinsic atomic orbital (IAO) and bond orbital implementation
- Improved performance in Boys localization
- Updated and improved mapspc program
- Atomic Mean Field (AMFI) spin-orbit coupling operators
- EPRNMR works with range separated hybrid functionals
- New molecular dynamics module by Martin Brehm (author of the TRAVIS
visualizer);