Hi, Two queries, answers to either one would be great:
1) I'd like to dump out statistics info on where most time is being spent at specified timesteps, instead of waiting till the end. I have shrink wrapped boundaries and a pretty large system so want to make sure I'm not frittering away cpu time on communication. I couldn't find anything in the documentation, is there a way to do this?
2) There is a warning when restarting from a restart file that was generated from a run with a different number of cpus. What are the major concerns here? I take it losing atoms is possible when first initialising?
1) I'd like to dump out statistics info on where most time is being
spent at specified timesteps, instead of waiting till the end. I have
shrink wrapped boundaries and a pretty large system so want to make
sure I'm not frittering away cpu time on communication. I couldn't
find anything in the documentation, is there a way to do this?
There are no commands to do this. You could add print statements
to the code for whatever timings you want. I usually just break a long
run into shorter runs, then you will see the stats at the end of each
short run.
2) There is a warning when restarting from a restart file that was
generated from a run with a different number of cpus. What are the
major concerns here? I take it losing atoms is possible when first
initialising?
The only issue is, depending on what commands you are using,
you may not reproduce the original run as closely as you might
hope. You shouldn't lose atoms. See the doc page for the
read_restart command for more details.