2NDORD Geometry optimization error:

[zozonmr]

Hi,

I am unable to carry out a simple 2NDORD or .NEWTON geometry optimisation with PCM model in DMSO because the analytical hessian is not available. I do know that this problem has already been posted once but following the instruction there did not solve my problem. The instructions were:

1. add nosymmetry to the *.mol file
2. remove vibration analysis and properties altogether.

However the problem persists. I have attached the output file.

Any idea what happened here? How can I tell dalton to calculate the analytical Hessian?

Thanks

[magnus]

It fails because there are no analytical molecular Hessian has not been implemented for B3LYP and btw also not for PCM as far as I can tell.

[kennethruud]

Hi!

I guess the previous example only involved a vibrational analysis only in the final step. The easy way out of this would be to remove the keyword ".2NDORD" under *OPTIMIZE. The default optimization scheme is a first-order scheme, and this will then be used.

I note your cavity have many warnings, and I would recommend adding

.AREATS
0.3

under *PCMCAV. I would also recommend not placing a sphere on each atom, but rather use a single sphere for a group of atom+hydrogens attached to it. Some examples of such an approach can be found in the literature in terms of radius size, and otherwise please consult the manual on how to input such a cavity.


Best regards,

Kenneth

[magnus]

I noticed that you used OpenMP threading which is not used in Dalton (although you can get some speedup if the linked math libs support it) so I would recommend to use MPI instead.

[zozonmr]

Hi,

Thanks for the prompt answers it is very helpful. I am not going to push 2ndord with b3lyp.

In a different pcm calc I put the spheres only on the heavy atoms I just want it to test this but thanks for the area tip I will set it 0.3.

The openMP is there since I think BLAS uses it but I may be wrong here. Is dalton running on GPU? I started exploring LSDALTON for geometry optimisation and then run the single point on dalton (for IR spectrum and NMR props) however I still have not found the optimal settings.

Thanks!

[magnus]

Even if your BLAS and LAPACK are threaded you will in most cases get much more from using MPI (except for CC calcs which aren't MPI parallelized), and as far as I know no part in Dalton is GPU parallelized, it only uses MPI parallelization.

[zozonmr]

Yes I have tried to compile with --mpi but I got an error message from cmake. Do I have to enable_language(Fortran) in the CmakeLists.txt file?
This the error on invoking ./setup --omp --mpi:
CMake Error at /usr/share/cmake/Modules/CMakeDetermineFortranCompiler.cmake:45 (message):
Could not find compiler set in environment variable FC:

mpif90.
Call Stack (most recent call first):
CMakeLists.txt:18 (project)


CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_Fortran_COMPILER_ENV_VAR
CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_Fortran_COMPILER
CMake Error: Could not find cmake module file:/home/takacs/DALTON-Source/build/CMakeFiles/2.8.11.2/CMakeFortranCompiler.cmake
CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_C_COMPILER_ENV_VAR
CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_C_COMPILER
CMake Error: Could not find cmake module file:/home/takacs/DALTON-Source/build/CMakeFiles/2.8.11.2/CMakeCCompiler.cmake
CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_CXX_COMPILER_ENV_VAR
CMake Error: Error required internal CMake variable not set, cmake may be not be built correctly.
Missing variable is:
CMAKE_CXX_COMPILER
CMake Error: Could not find cmake module file:/home/takacs/DALTON-Source/build/CMakeFiles/2.8.11.2/CMakeCXXCompiler.cmake
CMake Error: CMAKE_Fortran_COMPILER not set, after EnableLanguage
CMake Error: CMAKE_C_COMPILER not set, after EnableLanguage
CMake Error: CMAKE_CXX_COMPILER not set, after EnableLanguage
-- Configuring incomplete, errors occurred!

[magnus]

It seems that you do not have any MPI available. If this is on your own machine then you will typically have access to both OpenMPI and MPICH through the repository of your Linux distribution.

[zozonmr]

Hi,

This problem was solved completely and mpi is working like charm. The problem was that there were no links in /usr/lib to mpif90 they were in lib64. After that I have got speed now!

Thanks again!