Over the years, the kernel’s CPU scheduler has become increasingly aware of
how much load every task is putting on the system; this information is used
to make smarter task placement decisions. Sometimes, though, this logic
can go wrong, leading to a situation that Valentin Schneider describes as
“utilization inversion”. At the 2020 Power Management and Scheduling
in the Linux Kernel summit (OSPM), he described the problem and some
approaches that are being considered to address it.

Source: LWN.net – [$] Utilization inversion and proxy execution