- Efficiently develop new Mathematica codes and algorithms for parallel computing, combining
Mathematica's rich feature set with MPI.
- Run Mathematica symbolic manipulation across a cluster, as cluster computing handles more than numerics.
- Perform symbolic manipulation and processing that is inconvenient to do in Fortran and C, at a scale that only
clusters can handle.
- Use the environment of Mathematica to learn how to program clusters and supercomputers.
- Applications that take a long time (days or weeks to perform) on one node can be performed significantly faster (up to 8 times faster when utilized on a single 8-CPU node, or up to 32 times faster when utilized on a cluster of four 8-CPU nodes).
- Applications that require more RAM than one node can become possible to perform using all of the RAM in the cluster
(distributed memory model).