hi
I know it is possible to communicate between concurrent replica runs with lammps, e.g. PRD, but there is no standard way of obtaining variables from one replica on another. e.g. to correlate the stress or heat flux on one replica with another?
Any idea how hard this would be to implement?
thanks,
Reese
hi
I know it is possible to communicate between concurrent replica runs
with lammps, e.g. PRD, but there is no standard way of obtaining
variables from one replica on another. e.g. to correlate the stress or
heat flux on one replica with another?
reese,
can you be a bit more specific what it is that you are looking for?
i.e. which MPI tasks exactly, you want to have to communicate with
another.
Any idea how hard this would be to implement?
in principle, it would simply be a matter of generating a custom
communicator.for that. i've written/drafted code for such cases in the
past, so it would mostly be a matter digging up those old e-mails. but
before giving possible bad advice, i need to know what i need to dig
up.
thanks,
axel.
The temper command already does something
like this. It creates a communicator “roots”
that has the proc 0 of ecah replica, and uses
it to exchange temperature and potential
energy info between the replicas.
Steve