[ase-users] tip3p and QMMM

[Gianluca Levi]

Hi Daniele,
Most likely, the bottleneck is in the application of bond-length 
constraints to fix the geometry of TIP3P water. In ASE this is done with 
an implementation of RATTLE in constraints.py (FixBondLengths class). 
The implementation is currently not parallelized, and can be very slow 
for large numbers of MM molecules.
But, a developer is working at a C parallel implementation of both TIP3P 
and holonomic constraints, which should be available very soon to use in 
conjunction with QM/MM:
https://gitlab.com/juhanim/gpaw/tree/fast-holonomic-constraints
https://gitlab.com/gpaw/gpaw/merge_requests/313/diffs
Apparently, the implementation is at least two order of magnitude faster 
than the current implementation of RATTLE.
Best regards,
Gianluca
_______________________________________________________________
Gianluca Levi
PhD student
Technical University of Denmark, Kemitorvet Building 206, room 226, 2800 
Kgs. Lyngby
Mobile phone: 004550352010

On 17/01/2018 10:54, daniele selli via ase-users wrote:
> Dear ASE Users,
>
> I am trying to perform QMMM calculations with ASE.
> My MM part consists of only water molecules, so I would like to
> use the built-in calculator of ASE for the tip3p model of water.
>
> Everything works pretty fine for my tests, like a water dimer (one 
> molecule
> QM and one MM) or a small box of water (~100 water molecules, 4 of 
> which QM). The problem is that if I want to perform calculations on 
> much bigger systems (let say ~20k atoms with ~50 of them QM) I got 
> stacked.
>
> The calculation bottleneck surprisingly is the MM part. When ASE 
> calculate the interactions between the MM atoms and the ones between 
> MM and QM atoms it takes ages since it is not parallelized. So my 
> question is that:
> Is there a simple way to parallelize the tip3p part from ASE? I was 
> thinking to use some "Parallel" python, but there might be some 
> simpler solution.
>
> Thank you very much for your help,
> Cordially
>
> Daniele Selli
>
>
> _______________________________________________
> ase-users mailing list
> ase-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/ase-users

-- 
Gianluca Levi
PhD student
Technical University of Denmark, Kemitorvet Building 206, room 226, 2800 Kgs. Lyngby
Mobile phone: 004550352010



---
This email has been checked for viruses by Avast antivirus software.
https://www.avast.com/antivirus
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.fysik.dtu.dk/pipermail/ase-users/attachments/20180118/985d1342/attachment.html>