As far as I understand:
- CPUs can change their frequencies relatively quickly, which can be used for performance-power tradeoff based on load and priorities.
- big.LITTLE goes further – you can switch between powerful and energy-efficient cores in order to save more energy than underclocking allows you. Those cores vary by architecture, otherwise it would not make much sense. Usually, there is a mix of powerful cores with out-of-order execution and energy-efficient cores with in-order execution.
However, according to GSMArena, Nokia 7 plus has the following CPU:
4x2.2 GHz Kryo 260 & 4x1.8 GHz Kryo 260
That is, you have eight cores of the same type, but four of them can go up to 2.2 GHz, while other four of them can go up just to 1.8 GHz.
I believe we cannot use the same justification as for big.LITTLE. Instead of switching the cores, we could just scale the core's frequency.
Also, if the upper frequency is fixed (that is, if there are four physical cores that never go beyond 1.8 GHz), it implies uneven distribution of load and heat generation.