Chemistry Reference
In-Depth Information
Figure 25
Screenshot of some of the features available in the analysis code pizza.py.
write LAMMPS input/output, and convert to other formats, among other
things. Figure 25 shows a screenshot of several of these tools in action.
DL_POLY writes output to a series of files from which the user can extract
essential data fairly easily. It also has a Java-based graphical user interface
that helps with job submission and analysis. NAMD also writes out a sum-
mary of properties, and it has a seamless interface with VMD that enables
easy visualization and animation of trajectories. Calculation of other proper-
ties such as radial distribution functions, transport properties, or most
thermodynamic properties, however, generally requires that the user write
his or her own analysis software.
Operating Systems and Parallel Computing
Most large-scale atomistic simulations take place these days in a
cluster
computing
environment in which a large number of commodity processors
running Linux are hooked together to make a parallel computer. Of course,
there are also specialized high-performance supercomputers that are run at
scientific data centers. For the commodity clusters, parallelization is handled
with various implementations of the message passing interface (MPI). If you
work in a university or national lab environment, then there are most certainly
experts all around you who know all about this and can help with installing
