hi,
i have been able to compile exciting-code ( lithium) with gfortran
for serial and parallel (mpi only) runs. in parallel i can excitingmpi running
on 4 cores (system has one node with 32 cores). but the cpu
time seems to be more than for a serial run....is it possible that
the excitingmpi is making 4 copies of the serial runs.system has
Tue Oct 16 15:57:10 UTC 2012 x86_64 x86_64 x86_64 GNU/Linux
please suggest some ideas.
sushil auluck