The scheduler is a fundamental component of the operating system, though in a world where one user gets double digit cores, maybe not so critical as it was in the past where 100 users might be sharing a single unix system to do work.
If you have 20 jobs to run and a single core, how do you share the resource? Do you do a primitive everyone gets the cpu for 1/20th of a second routine? Context shifting is expensive. But letting one job run to completion can be upsetting to the other 19 jobs.
Lots of cores means a lot less waiting, but you want related jobs to be on the same core/cache part of the cpu so the latencies for memory fetches is lower.
And now this introduction of unequal cores makes the optimal scheduling even more difficult. The end result - the Intel 12/13 gen cpus look much more impressive on Windows 11. If you're on 10, I think the AMDs are still the clearer choice, esp when you add the power considerations in. And if you're on alternate OSs (vmware, for ex), you might be waiting for software updates, much like the Apple M1 users had been for a while.
