Define the terms Time sharing

Time sharing: The best way to understand the concept of time sharing is to contrast it with multiprogramming. Multipro­gramming allows multiple programs that are ready for execution to reside in primary memory. While one program is doing some­thing unrelated to CPU, such as input-output, another one can run on CPU doing computations. This way, I/O activity involving slower devices is overlapped by computations involving fast CPU in order to increase the utilization of resources. Typically, a program (defined as a process by the operating system) is sus­pended whenever it needs to do I/O, and the CPU is assigned to another program. After finishing I/O, the suspended program is again readied for CPU assignment.

Timesharing requires multiprogramming and not vice versa. In time sharing, in addition to multiprogramming where a process is suspended upon, say I/O request; each process (program) is regularly suspended after its time quantum (in milliseconds) expires. The purpose is to let multiple users and their programs to share the CPU by taking turns at quick time intervals (usu­ally in round robin fashion). Although the CPU can do only one thing at a time (e.g. can run one program at a time), the users feel as if they are simultaneously using the CPU. This idea is sometimes referred to as "virtual machines," where if there are ten users on a fast CPU, each feels as if he/she is using a machine ten times as slow (on the average, of course!).