Scaling issue while performing HPC simulation (transient multiphase CFD)
I am running transient simulations on a supercomputer cluster. I submit a job description that acquires 32 nodes, each with 16 processors. However I recently noticed that the total wall-time of the job reflects only 1 node being utilised, which suggests that 31 nodes are not being used. I have been informed that this is an issue of "scalability" of my simulation. I would appreciate some guidance on how to maximise efficiency. I'm also unsure whether this is an issue that has to be fixed in the fluent case file, or that I need to adjust the PBS scripts.