Does this fact that not all cpu cores / processing cores are equal anymore bring back a reason to implement the hardware scheduling mechanisms built into x86 and x64 architecture ? Or is there still a good argument not too use that and to stick with the software method>? (in some hypothetical new operating system)