Provide information how recent FDS versions perform when increasing number of processors

[saschagottfried]

For FDS 6.1.0 Thunderhead Engineering constituted that running the case with 6 cores (hyperthreading disabled) was the fastest run. The solution was computed almost twice as fast as with one processors.

How is that with newer FDS version? How adding more processors influences run times?

[saschagottfried]

Compute absolute and relative differences with regards to versions and OMP threads