{"id":1669,"date":"2023-11-22T15:21:25","date_gmt":"2023-11-22T14:21:25","guid":{"rendered":"http:\/\/oopm.org\/?page_id=1669"},"modified":"2024-11-26T14:31:57","modified_gmt":"2024-11-26T13:31:57","slug":"8-active-objects","status":"publish","type":"page","link":"https:\/\/oopm.org\/?page_id=1669","title":{"rendered":"13. Active objects and parallel programming"},"content":{"rendered":"
\"image_pdf\"<\/a>\"image_print\"<\/a><\/div>\n

A computer may execute many activities in parallel, which means they take place at the same time. This is possible since a computer may contain a number of processing units called CPU’s or cores. Each core may execute an activity. A core may also share its processing time between a number of activities in a way that leaves the impression that they execute in parallel – this is called time-slicing<\/em>. Activities executed in parallel are often referred to as parallel process<\/em>es and\/or concurrent processes<\/em>.<\/p>\n\n\n\n

A system of several computers connected in a network is another example of a system where activities are executed in parallel. This is referred to as distributed computing <\/em>and\/or distributed processes<\/em>.<\/p>\n\n\n\n

The writing of programs that execute just one activity is called sequential programming <\/em>and the activity generated by the program is called a sequential process.. <\/p>\n\n\n\n

In the real world there are many example of systems where several active entities\/agents work together on a common taks. This may be a group of people and machines building a bridge, a public office handling cases, … Such systems are complex in the sense that the entities have to coordinate their work and exchange information.<\/p>\n\n\n\n

The ability of a computer to execute activities in parallel has many advantages. It may reduce the time for computing a given task. By splitting a given task into subtask that can be executed in parallel, may speed up computation compared to execute them in some order. However, there is an overhead in splitting up a task since these tasks may have to coordinate their activities. This means that the execution time of the subtasks must be larger than the time to coordinate the tasks.<\/p>\n\n\n\n

The splitting of a task into subtasks may be the job of the programmers, but for some languages, the compiler behind the scene may split parts of a sequential program into subtasks that may be executed in parallel.<\/p>\n\n\n\n

Another reason to consider parallel programming is that the environment consists of a number of devices, like computers, servers, and I\/O-devices that all execute activities, and these cannot be replaced by a sequential program.<\/p>\n\n\n\n

We refer to the above situations\/cases for splitting a task into subtasks by the following terms:<\/p>\n\n\n\n