ABINIT Discussion Forums

Thank you so much for your suggestions!

Dear All:

When running a job in 2x12 cores, it works, however, when running a job iunder 4x12 cores, it gets stuck at

"chkinp: Checking input parameters for consistency", but there is no error report, and I'm well aware that I have 48 cores access (otherwise abinit wouldn't run )

Dear All:

In Abinit, is it possible to tune the Fermi Level or electronic occupation manually? If this can be done, the electronic doping, such as voltage gating, can be described.

Thanks a lot!

Dear All: I am calculating electron-phonon properties. The unit cell contains 5 atoms, with k-mesh 8x8x8, and q mesh 2x2x2. There are 8 in-equivalent q points. This leads to a 5x3x8=120 perturbations in MRGGKK input, and 8 DDB files in MRGDDB input. the mrgddb and mrggkk seems both OK. However, when...

Eventually, I used MVAPICH2, instead of openmpi, and it worked to the end of make!!! (7.4.3 tested, 7.2.2 not tested)

I'm not sure what is the origin of the problem, though.

Many thanks!!!

However, fortunately, when I do the simplest make, without 64bit flag or mpi, it works!!!

"--prefix=/home/ludwig/abinit-7.4.3 --with-dft-flavor=none"


Therefore, I think the problem comes from the mpi. I'll try another mpi and see.

One more thing worthwhile mention, that even if I compile as: ./configure --prefix=/home/ludwig/abinit-7.4.3 --enable-64bit-flags=yes --enable-mpi=yes --enable-mpi-io=yes --with-mpi-prefix="/opt/intel-openmpi/" --with-dft-flavor=none which is quite simple, the same error still exists durin...

This time I actually tried 7.2.2, but still have the same error: : catastrophic error: **Internal compiler error: segmentation violation signal raised** Please report this error along with the circumstances in which it occurred in a Software Problem Report. Note: File and line given may not be expli...

Thank you very much!!! I've attached the config.log here. As to the compiler version, the cluster admin says that despite it was version 12.1, they keep it updated regularly. I'm not sure what exactly that supposes to mean, but it seems the version could be newer. (which I don't understand since the...

I now understand, simply shut down the firewall/ virus software, and it works totally fine, and takes seconds.

The DoS uses several smearing technique, which may make it "over-complete", like coherent states. Thus I personally think it is quite OK if you normalize them using the total # of electrons in the system.

Dear All:

In the response function calculation, even running the tutorial (v 7.4.3), the MRGDDB takes infinite time, contrary to "a few seconds". Anything might go wrong? Thank you in advance!

I really don't think there is anything to do with the Abinit version or compiler version. Could you show your input file? Ah, I know what's your problem. The interatomic distance is 0.7880Bohr? Which is WRONG. I think it is likely to be doubled, or you used a wrong unit system. See here: m It says &...

Hell! Thank you for your prompt response. From the documentation, I can see that the version from "csupport" is : "Package ID: l_ccompxe_2011.11.339 l_ccompxe_12.1.5.339 This package contains: Intel(R) C++ Compiler XE 12.1 Update 5 for Linux* Intel(R) Debugger 12.1 Update 5 for Linux*...

Thank you so much!!!

Hi, usually the band structure can be plotted using the post-processing tools AbinitBandMaker.py, located /utils/users/ However after installing 7.4.3, I cannot find the utilities.. May I ask where they are? I cannot find them anywhere. Many thanks, Also, in this tutorial, it only says to use "...

I'm having the same problems, too!!! This is somewhat inconvenient since I presume we need the real space information and visualization...

Hi All:

Is there an elegant way to visualize the wave function in real space? The cut3d program restricts the results only at 1 k-points at a time, which is quite slow...

Many thanks!

There is no problem at all at the stage of configuration, and the make install is also totally OK if I do it in another cluster. However, the problem I am facing is : : catastrophic error: **Internal compiler error: segmentation violation signal raised** Please report this error along with the circu...

Well when I saw the original paper and see it is analytical form then I came to realize how it works.

Still there is one thing I am confused, that in a S-O coupling calculation, is hgh potential the only choice?

Since it deals with Spin-Orbit coupling, I was expecting a more complicated pseudopotential than the LDA-TM one, However, the HGH pseudo-potentials have only a few lines. Am I missing anything? What's more, for calculation of spin-orbit coupling in LDA, can I simply use LDA-TM? Or hgh is a requireme...

Thank you so much!! That indeed helps a lot!!

Dear All: I'm trying to calculate the Density of State of a system containing 27 Se atoms. The input file is attached as (Se_27PAW_atoms.in). It can actually run for a long while (cpu time > 20 hours), and even the Density file (_DEN) is obtained, too. However, after that it crashes. In the log file...

Dear All: Maybe it is a dumb question... But I'm eager to see the answer: when compiling Abinit, it needs #make multi_nprocs=N, when running Abinit, it needs mpiexec or mpirun, which needs to specify the number of cores, say M. Does the two have to be the same value (N==M)? Or they are simply irrele...

Hi, I don't know which category I should put this, but on this download website:

http://www.abinit.org/downloads/source- ... rch=Update

no matter which version I click, it goes into the same newest version.

So I'd like to report this problem. Thanks!