Probably different compilation flags or whatever. In fact it does not mean anything. Try to calculate energy differences between different systems and see if there is anything wrong. On Jun 5, 2014 11:00 PM, "xu yuehua" <xu11249...@gmail.com> wrote:
> > Hi : > > i want to compute the total energy for a fixed structure with siesta-2.0.2 > > and for some reason, i change the calculation from a cluster machine to > another, the input file are the same. > > but i found the total energy in the two machines is so different : > > one cluster machine: > > > siesta: Final energy (eV): > siesta: Kinetic = 150531.103378 > siesta: Hartree = 669505.403513 > siesta: Ext. field = 0.000000 > siesta: Exch.-corr. = -30265.217897 > siesta: Ion-electron =-1489097.505611 > siesta: Ion-ion = 547814.176449 > siesta: Ekinion = 0.000000 > siesta: Total = -151512.040167 > > > and the other: > > siesta: Final energy (eV): > siesta: Kinetic = 145945.644671 > siesta: Hartree = 665399.694935 > siesta: Ext. field = 0.000000 > siesta: Exch.-corr. = -29577.856052 > siesta: Ion-electron =-1479423.697556 > siesta: Ion-ion = 547814.176449 > siesta: Ekinion = 0.000000 > siesta: Total = -149842.037553 > > > so i am wondering which one is ok? > > why there is so huge difference? > > PS, the siesta code is compiled by the same intel compiler. > > i need your help! > > > > -- > lalala lalala >
