Main Page > Networked Inference Engine > Networking

Overview

The entire project is based on the premise that two brains are better than one. After working on our last project, the instant messenger bot, we realized that the problems we are tackling are simply too large for a single computer to handle efficiently in a timely manner. Hence, every computational exercise of any significance in the Networked Inference Engine will be distributed to a network of participating computers.

Whether this happens via a traditional server/client architecture, peer-to-peer, or something more exotic, our focus will be on utilizing as much processing power on as many computers on as many processors on as many cores as possible simultaneously. Our progress will be inspired by Google's MapReduce, although not necessarily analogous or even similar.