Multiprogramming systems

Concept

Multiprogramming keeps several jobs in memory so the CPU can switch to a ready job whenever the current job waits for input/output. Its central performance aim is to reduce CPU idle time and raise system throughput, which is the number of jobs completed per unit time. Throughput is different from the turnaround time of one particular job.

Application

1. With only one resident job, an input/output wait can leave the CPU idle even though the system still has useful work to finish.

2. With several resident jobs, the operating system can schedule a ready job during that waiting interval, so CPU work and another job’s input/output wait overlap.

3. Because a larger share of the same time interval is spent doing useful CPU work, the system can complete more jobs in that interval. Therefore the applicable statement is “Execute more jobs in the same time period.”

Contrast

• “Are easier to develop than single programming systems” concerns implementation effort; coordinating several jobs requires additional scheduling, memory-management, and protection mechanisms.

• “Execute each job faster” concerns individual turnaround time; multiprogramming improves aggregate resource use but does not guarantee that every job finishes sooner.

• “Are used in only large mainframe computers” imposes a hardware-size restriction that is not part of the definition of multiprogramming.

Cross-check and result

Compare two equal time windows. If the only resident job waits for input/output, some CPU slots are idle; if another ready job uses those slots, more useful processing is completed before the window ends. This confirms the throughput interpretation and the result “Execute more jobs in the same time period.”