19990709: t3e.cineca.it vs utsj90.univ.trieste.it

GOAL

Run exactly the same input on utsj90 (see 19990707) and t3e.cineca.it (present calc), in order to check reproducibility across different hardware (and also to setup things on t3e). (Note that a similar experiment performed previously which gave dramatically different results on the two machines is to be discarded because the CONFIG files were not identical (see 19990602)).

See also 19990728.

Results

dschsun1:8> pwd
/usr/people/balducci/tmp/dlpoly/t3e/19990709
dschsun1:10> for i in CONFIG.START CONTROL.START FIELD
>do diff -c $i ../19990707/$i
>done
*** CONTROL.START       Fri Jul  9 17:52:10 1999
-- ../19990707/CONTROL.START     Fri Jul  9 11:36:59 1999
***************
*** 66,72 ****
  # (e.g. terminating the current cycle)
  # When CPU_time>(job_time-close_time), the program will start closing
  # operations and will stop
! job time  2000
  close time  300
  
  finish
-- 66,72 --
  # (e.g. terminating the current cycle)
  # When CPU_time>(job_time-close_time), the program will start closing
  # operations and will stop
! job time  10700
  close time  300
  
  finish
dschsun1:11>

A remarkable difference in the two MSD plots is observed after $120\;ps$: can we archive it as due to poor statistics?