Computational Neighborhood, or CN, is a Java based research and teaching framework for distributed parallel computing. CN provides a compelling alternative to existing message passing frameworks like MPI by eliminating painful administration tasks and providing a simplified programming interface.

Running CN on machines that share a subnet creates an ad-hoc cluster or 'Neighborhood'. Nodes can come online and go offline as they please. Using a simple client API, you can write your own jobs for CN. As the programmer, you decompose the jobs into tasks, and these tasks are distributed across the Neighborhood equitably. The tasks can communicate with one another via message passing in order to complete their work.

To get started using CN, see GettingStarted.