The GAUSSIAN program system¶
A description of the quantum chemistry program system Gaussian
Gaussian is a computational chemistry software program system initially released in 1970. Gaussian has a rather low user threshold and a tidy user setup, which, together with a broad range of possibilities and a graphical user interface (gaussview), might explain its popularity in academic institutions.
Online info from vendor:¶
The support people in NOTUR, do not provide trouble shooting guides anymore, due to a national agreement that it is better for the community as a whole to add to the community info/knowledge pool where such is made available.
Additional online info about Gaussian on Stallo:¶
When publishing results obtained with the referred software referred, please do check the developers web page in order to find the correct citation(s).
The license of GAUSSIAN is commercial/proprietary.
The license of Gaussian contitutes of 4 site licenses for the 4 current host institutions of NOTUR installations; NTNU, UiB, UiO, UiT. In principle, only person from one of these instituions have access to the Gaussian Software system installed on Stallo.
- To get access to the code, you need to be in the gaussian group of users.
- To be in the gaussian group of users, you need either to be a member of the abovementioned institutions or provide proof of holding a license on your own.
Since Gaussian is a rather large and versatile program system with a range of different binaries, we would in general advice users to check whether their jobs are parallelized or not before submitting jobs. It would in general, unless every step is entirely well paralellized, allways be more efficient to split a complex many-step job into smaller paralell and serial parts/jobs so that also the overall utilization of hardware is improved. I you are in doubt whether or not your job will scale outside one node (=shared memory), go to the Gaussian application home folder and check if there is an *.exel version of the executable(s) you will be using. If yes, your job will generally work ok in parallell up to approx 300 cores (this is for the more advanced users).
We generally wants users to run as many nodes as possible to limit the walltime length of running jobs.
$ module avail gaussian
to see which versions of Gaussian are available. Use
$ module load Gaussian/<version> # i.e 09.d01
to get access to any given version of Gaussian.
The first time you run a Gaussian job?¶
Get the information you need here:
About the Gaussian version(s) installed on Stallo¶
Since the installs we have made on Stallo are somewhat different from installs of Gaussian elsewhere, we have gathered some information about this here:
Here we also adress issues related to running Gaussian in parallel, number of cores to use, memory allocation and special issues taken care of at install.