In HPC environments, scaling the computation over various nodes is handled by the job scheduling system and the configured MPI instance. So, you need to setup your desired parallelization (like the number of nodes and tasks per node) in the job file, and the scheduling system takes care of this. As a result, the syntax for running the Palabos code doesn't change from what you pointed out.
16:45 instead of
$ cd build; make; cd ..
on can use -C flag of make itself :
$ make -C bulid
and Thank you!
Thank you for pointing out this useful tip :)
Thank you for the guidance! @8:15 What would be the syntax for running this code on 4 nodes instead of 4 cores?
In HPC environments, scaling the computation over various nodes is handled by the job scheduling system and the configured MPI instance. So, you need to setup your desired parallelization (like the number of nodes and tasks per node) in the job file, and the scheduling system takes care of this. As a result, the syntax for running the Palabos code doesn't change from what you pointed out.
very much clear now. Thank you @@TuxRiders
@@idreeskhan-zp5ey no problem :)
Another question is could you please clarify how it captures the turbulence, whether any turbulent/LES model is implemented in this code?
Palabos supports turbulence. You can find more information about it in the software documentation and this paper: doi.org/10.1016/j.camwa.2020.03.022
Thank you for the guidance@@TuxRiders
sir can i get your mail id
you can find my contact details here: mbarzegary.github.io/contact/