Distributed computing can be used for many applications from mundane storage to tasks that that put a heavy workload on the central processing unit (CPU). Today's telecommunications network and the Internet itself are examples of ubiquitous distributed computing models. Each computer is autonomous yet contributes to the larger system, whether for communications and information, data processing, modeling or control systems.
Computer processing based on one computer forces all sets of data to be funneled through that computer's processor one set at a time. When there are large amounts of data to be processed, this can become time consuming because each set of data has to be resolved before the next one can be started. Distributed computing allows multiple pieces of a large data set to be processed simultaneously.
Information-sharing networks make heavy use of distributed computing. Today's telecommunications network and the Internet are effectively one, giant database. The information stored on all connected computers is handled autonomously but can be requested across the network by another resource.
Whether requesting a web page or a phone number, a member of a distributed network processes that request and sends the information back to the requester. This also applies to the concept of distributed backups. Server farms and datacenters make use of distributed computing to ensure redundancy in backups, so that all critical information is safe from the potential failure of one server within the network.
Distributed computing can also be used to process large quantities of information rapidly, breaking it into discrete portions that can then be recombined into the larger whole. This allows widespread data-set analysis. Other times, this can translate to direct input such as rendering farms wherein each frame of a computer-generated scene is broken down into parts that are then each handled by computers in the distributed cluster. The completed segments are then recombined into the whole.
Another use for distributed computing is for large-scale scientific modeling. Environmental models can have large numbers of variables that a single computer would have to work out one by one before embedding into a final model. Distributed computing allows each of those variables to be parceled out to other systems and allows the results to be generated much faster, in most cases in real time.
Industrial control systems as well as aircraft control systems make use of distributed computing in very direct ways. These clusters of computers oversee both types of systems in real time, constantly reporting results to each other as well as to human operators. Should there be a malfunction or breakdown in the industrial process, the network immediately can pinpoint where the malfunction is and route around it until it is repaired. In the same way, aircraft control systems can quickly figure out traffic patterns, trajectories and cleared runways for aircraft to safely and efficiently operate in airports, as well as route around trouble areas caused by weather interruption.