Increasing efficiency of SCF

Find answers or ask questions regarding Dalton calculations.
Please upload an output file showing the problem, if applicable.
(It is not necessary to upload input files, they can be found in the output file.)

Post Reply
Addiw7
Posts: 76
Joined: 05 Apr 2016, 17:47
First name(s): Dawid
Last name(s): Grabarek
Affiliation: Wroclaw University of Technology
Country: Poland

Increasing efficiency of SCF

Post by Addiw7 » 15 Aug 2018, 10:33

Dear All,

As I wrote in some previous posts, I want to perform calculations of one- and two-photon absorption spectra for large molecular clusters
(100-450 atoms).

The first problem I encounter is related to creepingly slow SCF procedure for CAM-B3LYP-D3BJ density even for relatively small
clusters of ca. 150 atoms. On the other hand, in some cases the SCF is faster for clusters of ca. 320 atoms but the calculations do not
converge as the energy gradient oscillates around 1.3e-4 Hartree and cannot reach the threshold. I think this is related to the
SCF algorithm used in Dalton. In contrast, the LSDalton seems more suitable as I get much faster SCF convergence, as suggested by prof.
Jensen in one of previous posts. However LSDalton lacks feature of polarizable or electrostatic embedding.

Thus my questions are:
1) how I could increase efficiency of Dalton for SCF calculations at DFT level of theory? I searched manual but I couldn't
figure anything out.

2) Is it possible to converge SCF with LSDalton and use the result for calculations of OPA and TPA spectra with Dalton?

Best wishes,
Dawid Grabarek

User avatar
magnus
Posts: 477
Joined: 27 Jun 2013, 16:32
First name(s): Jógvan Magnus
Middle name(s): Haugaard
Last name(s): Olsen
Affiliation: Hylleraas Centre, UiT The Arctic University of Norway
Country: Norway

Re: Increasing efficiency of SCF

Post by magnus » 20 Aug 2018, 21:01

Regarding 2), there has been some work to enable this feature but I'm not sure whether it is complete.

I'm not sure, but I think you can use ChemShell with LSDalton for electrostatic embedding. We are also working on polarizable embedding in LSDalton at the moment which will possibly be available in the next release or soon after.

Addiw7
Posts: 76
Joined: 05 Apr 2016, 17:47
First name(s): Dawid
Last name(s): Grabarek
Affiliation: Wroclaw University of Technology
Country: Poland

Re: Increasing efficiency of SCF

Post by Addiw7 » 03 Feb 2019, 17:52

Since Dalton 2018 was released, are there any new functionalities for more efficient SCF?
I read release notes so I see .FCKTRA for new transformation of two-electron integrals but I run tests and it
does not decrease computational time.

hjaaj
Posts: 307
Joined: 27 Jun 2013, 18:44
First name(s): Hans Jørgen
Middle name(s): Aagaard
Last name(s): Jensen
Affiliation: Universith of Southern Denmark
Country: Denmark

Re: Increasing efficiency of SCF

Post by hjaaj » 03 Feb 2019, 20:19

.FCKTRA is for transformation of 2-electron integrals to MO basis, it is NOT relevant for SCF (2-el. integrals in MO basis are not used for Hartree-Fock nor DFT). It is relevant for MP2, CI, MCSCF, ... It is useful to save disk space and for faster integral transformation in parallel.

Post Reply

Who is online

Users browsing this forum: No registered users and 6 guests