The
user application on its dedicated cores can effectively latch the
CPU in a spin-wait or sleep in a blocking idle state without
fear of another thread taking control.
This effectively reduces
dispatch latency to near zero, as there is never any resource
contention in terms of processor allocation. The benefit to this
approach is that the core allocation for kernel and user space
remains fixed, defined at the beginning