# SNIC & PRACE VASP Workshop

**Workshop webpage:**
https://www.nsc.liu.se/support/Events/VASP_workshop_2020/

**Quick guide for working on Tetralith:**
https://www.nsc.liu.se/support/Events/VASP_workshop_2020/quick_guide/

**URL of this document:**
https://hackmd.io/fM5moKdoT8idRQSWf1jD2Q

### Icebreaker: Please write your name, where you work, and optionally what you're working on

- Weine Olovsson: AE at NSC, working with materials research software, e.g. VASP, WIEN2k, QE, ...
- Peter Kjellström: Application expert at NSC.
- Diana Iusan: application expert at SNIC-UPPMAX, main field: computational physics; keywords: VASP, RSPt, EasyBuild, best practices, performance analysis
- Pritam Kumar Panda: PhD student at Uppsala University in the Dept. of Materials Theory Division. Specially, working on biomolecules adsorption on 2D materials.
- Yu Song: Postdoc at Department of Forest Ecology and Management, Swedish University of Agricultural Sciences. I am working on the geochemistry of mercury.
- Leon Engelbrecht: PhD student at the University of Cagliari, Sardinia (Italy). Mostly invloved in classical MDl simulations of liquids.
- Chandan De: Postdoctoral Researcher at IBS Center for Artificial Low Dimensional System, POSTECH, South Korea.
- Banasree Sadhukhan : Postdoctoral researcher at KTH, Stockholm. I am working on topology in magnetic materials.
- Sagar Sarkar : Post-Doc researcher at Asia Pacific Center for Theoretical Physics, POSTECH, Pohang, Korea.
- Pavlin Mitev: application expert, UPPMAX, computational chemistry
- Serzat Safaltin: PhD student & RA at Istanbul Technical University, dept. of metallurgical and materials eng. I'm working on interatomic potential development for ternary and quaternary metallic alloys.
- Anders Vesti: PhD at Lund University, mechanics, physics

## General Questions (day 1)
* Sagar Sarkar : What shall be the best value for NBANDS. I mean including more empty states in the calculation makes it more accurate (not sure though.. I have not checked). So shall we have to check for it like the convergence of it?
* https://www.vasp.at/wiki/index.php/Number_of_bands_NBANDS Yes, in principle one should check the convergency with respect to the number of bands. VASP will determine a default number, which should be sufficient in most cases. But for very accurate simulations, or for GW simulations for example, you may need to increase it. If you want to compare the total energies of different simulations, pls. keep in mind that NBANDS has to be the same in all these cases. Also, the default NBANDS will depend on how many cores you are running on, for this reason it is good to explicitly define it in INCAR.
Sagar: Thank you very much.
* Depending on parallelization (e.g. number of nodes, cores), VASP might change your NBANDS, so a good practice is to check with `grep NBANDS OUTCAR`

* Sagar : For Si dos plotting a very high k-mesh has been used 21x21x21. How to judge that correctly. I mean if I know the converged k-mesh for example say 17x17x17 for this system, then what will be the mesh size for the corresponding dos calculation.
* It depends how smooth you'd like the curves to be. Roughly double the scf one may be a good starting point. The BZ integration method and SIGMA will also influence the smearing. Some recommendations from the VASP developers: https://cms.mpi.univie.ac.at/vasp/vasp/Accurate_DOS_Band_structure_calculations.html
Sagar : Okay. Thank you very much.

## Hands-on Questions (day 1)
* Anders: When we use ASE to find the volume of one atom, how do we know that the volume is the shape of a cube and not a sphere? E.g. we find a=(4*V0)^1/3.
* The shape of the fcc unit cell is cubic, so the volume is a^3 with each side a. An fcc unit cell contains 4 atoms. From VASP we got the volume V0 for one atom. Also compare with what you get from `grep volume OUTCAR` and `grep volume vasprun.xml`

* Sagar: For partial dos plot in p4vasp, what is the setting First band = 0, Last band = -1 ?
* Maybe "-1" is a way to say "all" in p4vasp

* Sagar: In p4vasp, where is my Fermi energy level?? Specially when I plot the band structure?? Is it at 0.0 shifted or the Fermi energy printed at the OUTCAR file?
* It's automatically set to 0.0 in p4vasp, but in the VASP output it's not set, so you would need to check its value
okay thank you.

* Sagar : Then my band structure for Si says it is gapless / metalic. I expect at least a wrong band gap.
* fcc Si is metallic, but cubic diamond Si has band gap.
okay thank you.

* Sagar : Some points regarding the band structure plotting of Si. When I compare with the band structure given in VASP Wiki, qualitatively it matches with it but there are quantitative differences. (a) There are more bands in my plot. (b) The Fermi energy is wrong. I need to substract the fermi energy of the SCF calculation to make it equivallent to the VASP Wiki example, but still the Fermi energy do not match. However I am using a lattice parameter 3.875 compared to 3.90 used in the Wiki.
* Wrong depends what we compare with, so we'd need to do the calculation exactly the same as was done in the VASP wiki, so the same lattice parameter and settings in INCAR.
okay thanks
* Sagar : May be a small error. For DOS calculation of diamond Si, the INCAR suggests ICHARG = 2 in place of ICHARG = 11
* It's for a single step calculation for DOS in one go. For a 2nd step calc. it's correct that one would set ICHARG=11 and copy CHGCAR from previous calc.
okay thanks

## General Questions (day 2)
*

## Hands-on Questions (day 2)
* Sagar : Using ISIF = 3 I get the optimized lattice parameter and cell volume very close to the VASP Wiki report. I have 2 questions:
* a) It was suggested to rerun the calculation using the final CONTCAR and check if there is any change. I foung no change, but why we need * to check that?
Answer: You need to copy your CONTCAR to POSCAR and re-run the simulation - just a simple self-consistent run. This step is preferred if you want more accurate total energies.
* Comment Sagar: Thanks .. this I agree. But I was asking about the extra task that says : "Repeat the volume relaxation calculation in a new folder, but this time start from the relaxed structure CONTCAR, by copying it to POSCAR. Does the structure change, if so, how much?"
Answer: depends how complicated your structure is. If you have a monoclinic structures, it may. Allow me to rephrase: whenever the cell shape changes a lot, it is good to do this step.
* b) Though the volume changed the shape of the cell do not change. What happens for a non-cubic cell? where we can change a,b,c and the angles to change the volume. So, which one is better? EOS method where shape remains fixed or the ISIF = 3 where all changes.
Answer: ISIF=3 is faster for the user in the sense that VASP will optimize the structure considering all degress of freedom (volume/shape/internal coordinates). I prefer to keep the volume fixed and optimize the positions and the cell shape, if needed, only.
* Comment Sagar : Okay, Thanks a lot. So, I can do an EOS method to get the equillibrium volume and then do the ionic position and cell shape optimization for that volume. So, this would be a reasonable approach ??
Answer: You should allow the shape and positions to change for different volumes, as otherwise you may not find the equilibrium structure.
* Comment Sagar : Okay thanks a lot. I understand. So this suggests that I shall take the Structure (that I get relaxing shape plus ion positions for each volume) corresponding to the volume that gives the minimum energy. But then, with this data, if I do a EOS, it will give me an equillibrium volume but that volume has no structure associated with it. My question is shall I have to do the EOS fitting now or just take the minimum energy structure?
* Sagar : About ISIF = 2 with EDIFFG. EDIFFG can be positive or negative having different meaning. So, which one is preferrable? or do we need to use both for different purposes?
Answer: "If the change in the total (free) energy is smaller than EDIFFG between two ionic steps relaxation will be stopped. If EDIFFG is negative it has a different meaning: In this case the relaxation will stop if all forces are smaller than |EDIFFG| . This is usually a more convenient setting." The negative EDIFFG is typically preferred.
Sagar: Okay .. thanks
* How to automatically generate the KPOINTS file?
Answer: https://www.vasp.at/wiki/index.php/KSPACING

## Notes:
Feel free to share any useful links.
- Useful to check when considering structural relaxations in VASP: https://www.vasp.at/wiki/index.php/Energy_vs_volume_Volume_relaxations_and_Pulay_stress
- Another useful database of crystallographic structures is the ICSD one: https://icsd.fiz-karlsruhe.de/
- VASP examples: https://www.vasp.at/wiki/index.php/Category:Examples
- Boost tools at NSC to increase priority of important jobs or prolong walltime: https://www.nsc.liu.se/support/batch-jobs/boost-tools/
- Diana: some of my observations on the VASP parallelization flags: `KPAR` and `NCORE` are the parallelization flags that influence the speed and memory requirements of a job the most. It is a good idea to start optimizing these for your simulation and then move to optimizing `NSIM`. The memory requirements increase linearly with KPAR, but not as drastically as decreasing NCORE. The number of cores of the simulation has to be divisible by NCORE*KPAR. [NCORE - how many cores work on one orbital]
*The optimal NCORE, KPAR, and NSIM will depend a lot on the system you are trying to simulate, the number of k-points used, but also on the architecture of the cluster, the memory available on the node and how many cores or nodes you'd like to use.*
You may read further here: https://cms.mpi.univie.ac.at/vasp/vasp/Parallelisation_NPAR_NCORE_LPLANE_KPAR_tag.html, but please keep in mind that this guide is very outdated. I always recommend you do a few tests (with a few), especially when starting out with a new system. You are going to save time on the long run.
- VASP calculation for organic polymers: https://pubs.rsc.org/en/content/articlelanding/2014/cp/c4cp00146j#!divAbstract
- ...

---
###### tags: `workshop` `VASP`