Austin Zadoks
Austin Zadoks
It looks like you're doing a bit of sanity checking in code (e.g. that the smearing type is known). This makes sense if the AiiDA plugin writes a `.jl` file...
The general idea of AiiDA is to completely separate AiiDA itself and the calculation being run. As such, AiiDA is run on some host computer; for the sake of simplicity...
I wasn't sure if they account for the half-weight at zero. I did a bit of testing comparing the dot product and the ctrap with the half-weight, and they seem...
It looks like ~78% of time in initializing the `PlaneWaveBasis` for small Ecut and few k-points with UPFs is taken up by `_besselj`, which is a call to Amos. ...
Ok sweet, I did write up the hand-coded spherical Bessels, which was already much faster. I thought that `LoopVectorization` wasn't a dependency of DFTK, do I didn't give it a...
```julia # For Silicon, 2-atom cell, si-q4.upf Ecut = 80 kgrid = [12, 12, 12] julia> basis = @benchmark PlaneWaveBasis(model; Ecut=Ecut, kgrid=kgrid) # benchmark BenchmarkTools.Trial: 1 sample with 1 evaluation....
I think this is as you suggest(?), but it seems to break `@turbo`. ```julia @inline function sphericalbesselj(order::Val{0}, x) sin(x) / x end @inline function sphericalbesselj(order::Val{1}, x) (sin(x) / x -...
It seems like `@simd` also doesn't work with the `Val{l}` dispatch. With `@simd`: ```julia # For Silicon, 2-atom cell, si-q4.upf Ecut = 80 kgrid = [12, 12, 12] julia> basis...
> Just out of curiosity I tried a UPF for a pretty big atom (Hafnium) and got greeted with NaNs. I have not investigated further as I think we should...
> Agreed we should not have these in the repo. Gth are fine because they are super small fine, but beyond that it's fine to require people to get the...