Looking on github I see that there are some packages in Julia which are tagged with density functional theory (DFT). I wonder, if any of them can be used for calculating band-structure of graphene (for example figure 3 in this article ). PS. I never have calculated band structures before so my under…

The simplest and most reliable way to get you DFT from Julia is via PyCall and ASE. I’m maintaining a simple interface ASE.jl if this is of any use. From ASE you can of course link any DFT code. GPAW seems particularly easy to use. I wouldn’t mind expanding the ASE.jl interface to provide convenienc…

If you are not familiar with bandstructure calculation with DFT, I think it is good to start with tight binding method (especially for 2d materials) as others have been mentioned before. You might want to port http://physics.rutgers.edu/pythtb/ to Julia as an exercise (oh, and compare the performanc…

Here is what I get using PWDFT [TEMP_GRAPHENE] Using 15 Ha cutoff and 5 bands (2 atoms per unit cell plus 1 empty state)

Looks amazing :slight_smile: I need to get two additional levels from the top and add KH path. Would you mind to share the code you used? Documentation seems to be lacking at this point.

[image] f-fathurrahman: DFT calculation is rather resource-consuming and many packages are written in Fortran. Which implicitely raises the question about parallelization. Briefly looking into PWDFT.jl there is no parallelization yet ? What would be the approach with julia ? Does MPI.jl work …

My understanding is that PW codes rely essentially on parallel FFT.

[image] cortner: My understanding is that PW codes rely essentially on parallel FFT. Am not sure what to make of that remark. FFT primarily for the Poisson solver (or GW if you want to do GW), that is certainly true. But that is usually done with MPI (because usually you overflow a single no…

[image] ckoe-bccms: FFT primarily for the Poisson solver (or GW if you want to do GW), that is certainly true. But that is usually done with MPI (because usually you overflow a single nodes memory at some point even ignoring cores per node and speed) by distributing explicitely one fft dimension…

[image] stevengj: If you look at the MPI section of the FFTW manual , you’ll see that the caller is required to explicitly create distributed arrays. I have to admit that I never had to think about MPI fftw and its api. I only know for sure that vasp does not use it (licensing reasons? , altho…

[image] ckoe-bccms: Yes. My point was that this is not fft and still can be time and communication expensive, so must not be forgotten even with plane-waves. Well, google gives these slides for pwscf which show on page 27 (Broadwell row) if I understand them correctly that diagonalization can…

Calculating bandstructure of graphene

Specific Domains Modelling & Simulations

antoine-levitt August 22, 2018, 8:55am 18

I should mention @John_Gibson had similar concerns in Implementing parallel fluids code in Julia (parallel FFTW, MPI.jl, ...) - #34 by favba

Topic		Replies	Views
How to improve the scaling of Julia code aimed at multi-node parallelization? Julia at Scale linearalgebra , distributed	38	1011	August 14, 2024
Tutorials or examples on multithreading and distributed processing for numerical PDEs New to Julia	8	1374	June 2, 2021
Implementing parallel fluids code in Julia (parallel FFTW, MPI.jl, ...) Numerics	34	10268	September 21, 2017
How to convert a thread-parallelized code into a core-parallelized code? Julia at Scale multithreading , linearalgebra , distributed , threads , matrix	3	372	May 19, 2024
Distributed.jl vs MPI.jl Performance question , package , mpi , distributed	26	7167	January 31, 2022

Calculating bandstructure of graphene

Related topics