Hardware reporting now covers all nodes

Hardware reporting now covers all nodes

The CPU and GPU hardware reporting and checking were only printed
for the node(s) of the master rank(s) of each simulation.
Now the most important CPU and GPU information is reduced over
MPI_COMM_WORLD. These results are printed and checks/warnings are
now based on this.
Because this is now system wide, the printing is moved up to directly
after the detection and GPU info is printed together with CPU info.
The Gromacs reference printing is moved to after hardware printing.

Refs #1643.

Change-Id: I974871c9a293a922dd1db9dcf633d485f62996ae