计算化学公社

标题: Question regarding Gaussian Linda parallel setup [打印本页]

作者
Author:
aitouna    时间: 2020-1-16 03:11
标题: Question regarding Gaussian Linda parallel setup
Dear Professor Tian Lu,
I read your article "Gaussian03's Linda Parallel Setup with VMware Virtual Cluster http://sobereva.com/47", about Setting up TCP-Linda, and i would like to know if it's possible to use the linda 7.1 that you used in the article with other version of gaussian like gaussian 09  D.01.


Sincerely,



作者
Author:
zjxitcc    时间: 2020-1-20 21:43
本帖最后由 zjxitcc 于 2020-1-20 21:47 编辑

There is a Linda Manual on Gaussian website http://gaussian.com/g16/lindamanual.pdf.gz

Few people have Gaussian with Linda (because it's expensive). If you still have questions, you may get quick response if you ask the Gaussian Company (www.gaussian.com) for help (e.g. via email). You (or your lab) bought this software from their company, and their are supposed to answer your question.

Before you get any answer, what you need to do is: just try. See if it works.


作者
Author:
sobereva    时间: 2020-1-21 01:04
Practically speaking, using Gaussian with Linda is never a good idea, because not only the Linda version is much more expensive, but also the performance loss due to cross-node computing is usually significant.
作者
Author:
aitouna    时间: 2020-1-21 02:20
Dear Professor Tian Lu,
I spent a lot of time searching for methods to make the calculation more faster, and i always find that the cluster is the best one, and to use gaussian with the the cluster we need TCP-linda.
If there is a good alternative to speed up my calculation, i would like you to advise me.
Note : i only use gaussian software.
作者
Author:
zjxitcc    时间: 2020-1-21 10:35
You have many alternatives, i.e. many free softwares. E.g.

ORCA (free, not open source, can be downloaded at https://orcaforum.kofo.mpg.de/app.php/portal)

PySCF (free and open source, can be downloaded on GITHUB. If you want to run it on more than one node, you need its mpi version.)

GAMESS(free and open source, but may be not user-friendly to green hand)
作者
Author:
aitouna    时间: 2020-1-21 18:49
Dear zjxitcc,
Thank you very much for the information, i appreciate your help.
me and my colleagues we are not familiar with the softwares you mentioned, we always use gaussian.
we left a lot of topics because we know it will take a very long time of calculation, and that is very annuying.
作者
Author:
aitouna    时间: 2020-1-21 19:00
Dear All,
Please, I would like very much your kind advice about the suitable hardwares to speed up my calculation using Gaussian software.


Sincerely,
作者
Author:
sobereva    时间: 2020-1-24 17:50
There is no effective advice to speed up your Gaussian calculation without any detailed information about you input files. At least, using Linda is not a good choice, rare people use Linda version of Gaussian.
Using ORCA in single point calculation to replace with Gaussian is an absolutely good idea. Even if you are not familiar with it, if you plan to engage in quantum chemistry calculations for a long time, this code must be learned (it is not difficult to use).




欢迎光临 计算化学公社 (http://bbs.keinsci.com/) Powered by Discuz! X3.3