In computer terminology, parallel processing is the simultaneous processing activities of multiple processors being attached to a single computer system or in a network. A program or set of instructions are then divided among these processors and they simultaneously start working on the same program at once. This technology is especially useful in complex programs where the time complexity is more than n3. Thus if multiple processors work on the same program, it takes less time. These multiple processors are some time in one computer and such computer are then super computers.

The number of processors for a system should also be in a specific limit because if it already has several dozens of processors then adding another processor will improve the performance too small and the expenses will also be out of control.

Parallel processing has one concept of multiprocessing and in such systems two or more processors share the work of the execution. In earlier versions there was a concept of master and slave. One processor would be the master of all other slave processors. Then another technology of symmetric multiprocessing systems gave a new style. Here all the processors are equally responsible for the flow of data and the execution decision.

Parallel processing is nowadays used in many organizations like weather forecasting etc where there is a need of fast execution at less time. They cost much but give a benefit of time. If an organization decides to have parallel processing system in their work area, they should first examine their priorities. If they can’t compromise with the expenses then having two three processors will be ok than having 20 or 30 processors at high cost.

In such systems the data and material is passes through a communication medium if they are not sharing a same memory. Specially in a network at an organization, the program’s modules need a constant communication if they are distributed among multiple processors.