ABINIT Discussion Forums

Hi Joe, it's good to see you again. can someone provide a rough estimate for the memory needed per processor during the screening calculation and the BSE calculation It's possible to obtain an estimate of the memory requirements by using max_ncpus > 0 in the input file. The code will print in the ma...

runtests.py cannot find the abinit executable in /home/nissar/Documents/Program/abinit-8.0.8/src/98_main/abinit If you have build the code in a different directory e.g. in /home/nissar/Documents/Program/abinit-8.0.8/tmp cd to tmp/tests with cd /home/nissar/Documents/Program/abinit-8.0.8/tmp/tests an...

Hi Ryan, I'm about to embark on installation on various computers, e.g. Stampede, Maverick, Jetstream, Comet, Gordon, OSG, Bridges, etc. But every computer will have its own challenges, so it seems like it would be sweet if the abinit community shared config/install files and notes on each of these,...

Could you please provide an input file?

Hi Marco, I have access to another cray machine and we have compiled abinit with PrgEnv-intel and PrgEnv-gnu. Note that ftn compiles binaries that are supposed to be executed on the compute node so I would suggest to request an interactive session with salloc (I assume the machine uses Slurm) and tr...

Use:

--with-psml-incs="-I/directory/with_pspml_include_files/"

--with-psml-libs="-L/directory/with_pspml_libs -lname_of_library"

If the configure fails, send us the config.log file

It's not perfectly clear to me what you want to compute. You can use abinit and the cut3d post-processing tool to export the Bloch states and the crystalline density obtained at the end of the SCF cycle in a format that can be read by visualization software such as Vesta Xcrysden etc. This will allo...

I imagine you are using abinit7. In this case, you have to set the value of extension_switch to 0. Just open the file in an editor and change the line with extension_switch. From the web-site: Version Compatibility The PDv0.2 table is directly compatible with ABINIT version 7.10.5. For previous ABIN...

I'm attaching another version of the script that seems to work.
I had to change the file extension by adding a `.log` at the end in order to upload the file.

HTH
M

I wonder if this is due to the script cannot cope with the change of file extension introduced in abinit-8.0.7, i.e. from _HIST to _HIST.nc, or other issues? Yes, could you try this patch? === modified file 'doc/tutorial/lesson_paral_moldyn/diag_moldyn.py' --- doc/tutorial/lesson_paral_moldyn/diag_...

There's been some work to interface Abinit8.0.7 with Wannier2 but non-collinear magnetism
(nspinor = 2, required for SOC calculations) is not supported yet in this part of the code.

To fix this problem in version 8.0.6, one has to change the file ~abinit/tests/__init__.py by commenting the string "bigdft_paral" at line 84. See diff below. === modified file 'tests/__init__.py' --- tests/__init__.py 2016-04-23 10:05:44 +0000 +++ tests/__init__.py 2016-05-30 22:34:57 +00...

Possible explanation: Your Abinit is linked against the Blas implementation provided by Mac OSx (a.k.a. veclib) The ABI of the veclib routines zdotc and cdotc is not compatible with Fortran and this causes nasty segmentation faults at the beginning of the run. Were this true, I would recommend passi...

The configuration exited successfully as one can see from the last portion of config.log configure: exit 0 The warnings you are seeing are due to the fact that configure is "testing" your system and some tests are expected to fail. Have you already tried to compile libxc with `make`? Does ...

The PAW+UPF files are not readable by abinit. We support UPF1 for norm-conserving pseudos but PAW+UPF is more difficult to handle. The internal implementation of the PAW formalism in Abinit differs from the one assumed in the PAW+UPF format and it's still not clear if one can convert the UPF data in...

We have observed similar problems when running structural optimizations on intel architectures with recent intel compilers. The problem seems to be related to the optimization (vectorization) performed by the compiler that leads to erratic/non-deterministic behaviour on the different MPI procs. Ther...

However, the real error in the calculation was not the pseudopotential file, but was an incorrectly set input variable nspinor. If possible, it would be nice to catch this error when Abinit reads in the pseudopotential file by comparing the value of nspinor and the extension switch. If they are the...

Replace src/52_fft_mpi_noabirule/dfti_fftrisc.finc with the file attached. btw: The DFTI library provided by MKL should be activated only if you are compiling Abint with OpenMP threads (and indeed you are doing). Remember to pass --enable-openmp to the configure script because there are sections of ...

Does the problem disappear if you disable bigdft with:

with_dft_flavor="atompaw+libxc+wannier90"

?

with_linalg_libs="-L/home/xxx/lib/magma/lib -lmagma -L/opt/intel/mkl/lib -lmkl_scalapack_ilp64 -lmkl_intel_ilp64 -lmkl_core -lmkl_intel_thread lmkl_blacs_intelmpi_ilp64 -lpthread -lm -ldl -liomp5 -lmkl_lapack95_ilp64" Use the lp64 version of the mkl library instead of ilp64. The ilp64 ver...

But I don't know why the GPU version work only when the npfft equal to 1. Because FFTs in parallel with MPI and multiple GPUs are not supported. For the time being, one has to use npfft = 1 in the input file. And I want to ask you whether there are effects on the output result when I change the npf...

I will have a look at it.
Note, however, that for the time being it would suggest to use norm-conserving pseudos for self-consistent GW calculations.
Recent papers showed that GW with PAW requires specialised pseudopotentials in order to get accurate results.

Which value of occopt are you using? Have you looked at the occupation factors obtained by Abinit at the end of the GS calculation? Note that in metallic systems, excitonic effects are strongly dumped and one can obtain reasonably good results by simply using the Random-Phase approximation. The BSE ...

Use nsppol=2 and nspden = 2 everywhere.
The WFK file must be regenerated with these input variables.

Hi, You get an error message in the log file only if the the MPI process that encounters a problem is connected to the log file. Usually the MPI proc with rank == 0 in MPI_COMM_WORLD is connected to std_out, while the other processes are redirected to dev_null. In version 8, the code produces the__A...