In WWDC, when Craig introduced how they peak CPU usage "instantly" peaks in iOS12, to give you better scrolling (faster performance), does that also mean it does not time slice between apps anymore and instead slows down all other running apps just to draw performance attention to the front-most only while its open?
Basically having the total reverse of how Apple previously did it.
Having better performance is always good, but it looks like it now no longer if across all running apps..
Only the front most app, get all the performance increase. So, its no longer balanced.. Compare that to IOS 11, where CPU was time slicing.
It doesn’t just apply to scrolling, though that was the example given. The idea is that the kernel scheduler knows what requires CPU resource in user space as it occurs. For the iPhone that’s the GUI of the foremost app, plus any of the types of background threads. For the iPad that’s up to three GUIs, plus any background threads. Each and all of them will be serviced more quickly by a kernel that’s now more responsive to demand, rather than responding to demand some time after the fact.
Last edited: