[gmx-users] compiler options

[Shrivastava, Indira (NIH/NCI)]

 Hi Gromacs users,

 I am doing some benchmarking for a dual processor xeon cluster. 
 I ran a 500 ps simulation on two xeons, with identical configuration 
 and compiled gromacs software, but I got different trajectories (as 
seen in the rmsd's of the carbon alpha atoms from the two trajectories).
 There was a difference of ~0.05 angstroms in the rmsd's at each time step. 
Can someone tell me why there is this difference in the trajectories? 

 thanks

 Indira