“…In this example, it is necessary to take into account latencies due to the cyclic dependencies, not shown in the listing, between tasks, i.e., those due to the task reading the variable i and the task which decrements it on the one hand, and the reading/writing of variables M on the other hand. Finally, P-WARE [22,23,25] is used to evaluate the performance of global scheduling of the system (rate, bandwidth, execution time, etc.). A sequential scheduling has an execution time of six cycles and give a rate of one output every six cycles, whereas pipeline scheduling on two processors P1 and P2 produces a better rate for the same execution time as shown in Table 6.…”