RTD "getting started" suggestions

I'm going through the "Getting Started" section of the RTD. I've noticed some things that in my opinion could be improved.

1. There are some formatting issues:

E.g. In Compiling SWIFT/Optional Dependencies, we have

There are also the following _optional_ dependencies.

I think _optional_ is intended to be italic, not have enclosing underlines.

In Command line options, the last sentence is

See the file examples/parameter_example.yml for an example of parameter file.

the file name should have inline code format.

2. In the section "What about MPI? Running SWIFT on more than one node", we write

You will need some GADGET-2 HDF5 initial conditions to run SWIFT, as well as a compatible yaml parameter file.

Three points here:

• I think it would be better to call it "initial conditions in the GADGET-2 HDF5 file format"
• We should add a reference to what that format is. Either to somewhere in the docs itself, or to some external reference. Does anyone have a reference where the GADGET-2 file format is defined?
• Same for yaml parameter file: We should cross-reference this to the corresponding section of the docs.

3. Section "Running on Large Systems":

We don't mention hyperthreading. According to a recent discussion on slack, hyperthreading is recommended. We should note that somewhere. Secondly, I think the choice to use both 16 nodes and 16 cores per chip is a bit unfortunate. To make things clear, I'd change the number of nodes. E.g.

Your batch script should look something like the following (to run on 8 nodes each with 2x18 core processors for a total of 288 cores):

#SBATCH -N 8  # Number of nodes to run on
#SBATCH --tasks-per-node=2  # This system has 2 chips per node

mpirun -np 16 swift_mpi --threads=18 --pin parameter.yml

And add instructions for hyperthreading:

It is also recommended to run with hyperthreading enabled, if your system and hardware allows it. Your batch script should look something like the following (to run on 8 nodes each with 2x18 core processors for a total of 288 cores, with hyperthreading)

#SBATCH -N 8  # Number of nodes to run on
#SBATCH --tasks-per-node=2  # This system has 2 chips per node

mpirun -np 16 swift_mpi --threads=36 --pin parameter.yml

I'd be willing to adapt the changes and push a MR, but I wanted to check in with you for approval first. @matthieu , @pdraper , @jborrow , any thoughts?