-In the remaining examples, it's assumed that memory access isn't the dominant consideration. Unless otherwise stated, that all threads receive comparable computational resources. In these cases, the choice of schedule for a `for` construct depends on all the shared work that's to be performed between the nearest preceding barrier and either the implied closing barrier or the nearest upcoming barrier, if there's a `nowait` clause. For each kind of schedule, a short example shows how that schedule kind is likely to be the best choice. A brief discussion follows each example.
0 commit comments