A system is said to be hard real-time if the correctness of an operation depends not only upon the logical correctness of the operation but also upon the time at which it is performed. An operation performed after the deadline is, by definition, incorrect, and usually has no value. In a soft real-time system the value of an operation declines steadily after the deadline expires.

Hard real-time systems

Hard real-time systems are used when it is imperative that an event is reacted to within a strict deadline. Usually such strong guarantees are required of systems for which not reacting in a certain window of time would cause great loss in some manner, such as physically damaging the surroundings or threatening human lives; although the strict definition, is simply that missing the deadline constitutes complete failure of the system. Systems that always have hard real-time constraints (due to the potentially severe outcome of missing a deadline) include nuclear power stations and car airbags. In the context of multitasking systems the scheduling policy is normally priority driven pre-emptive schedulers. Other scheduling algorithms include Earliest Deadline First, which, ignoring the overhead of context switching, is sufficient for system loads of less than 100%.

New overlay scheduling systems, such as an Adaptive Partition Scheduler assist in managing large systems with a mixture of hard real-time and non real-time applications.

Real-time computing