Preliminary results for parallel implementation of the DEM algorithm
on a MIMD distributed-memory type architecture are modest and the
characteristics of the performance are considered to be dependent on
a number of factors: number of particles, number of boxes (internal
gridding), geometry of the compacted particles, mechanical of the
particles, and the frequency of input and output. We expect to
formulate this dependency so that more improvements can
be done to the speedup and efficiency of the parallel code.
Also, test cases with larger number of particles will be performed
on a higher number of processors.