MPIRUN

can any one clear my doubt.

the different number of processors used in mpirun will it give the different result for the same simulation inputs on same machine.

For example i ran the ploymer chain simulation which is given in the Lammps tutorial site. i had simulated in the same machine for two times as follows

  1. mpirun -np 4 lmp_linux -in in.*

  2. mpirun -np 32 lmp_linux -in in.*

but when i check the log file i am getting totally different values.

Is it the error or what it is?

check the attached log files.

log.PE_cl100.txt (9.19 KB)

log.PE_cl100 for 4 proc.txt (14.5 KB)

can any one clear my doubt.

the different number of processors used in mpirun will it give the
different result for the same simulation inputs on same machine.

i already answered that question. the behavior that you observe is correct.

axel.

100 atoms on 32 procs - extreme!

And even on 4.

Oleg.

21.09.2013, 12:35, "Thamarai Kannan" <[email protected]...>: