Why might a heterogeneous or asymmetric chip-multiprocessor be preferable to a homogeneous or symmetric one?

How do pipelines that support in-order and out-of-order execution differ in their microarchitectural components?
March 21, 2023
How do superblock and trace scheduling differ?
March 21, 2023

Why might a heterogeneous or asymmetric chip-multiprocessor be preferable to a homogeneous or symmetric one?

COMPUTER SCIENCE TRIPOS Part II – 2013 – Paper 7
Comparative Architectures (RDM)
(a) Why might a heterogeneous or asymmetric chip-multiprocessor be preferable to
a homogeneous or symmetric one? [5 marks]
(b) You are a computer architect working on the design of a new processor for
the mobile phone market. An initial analysis of applications suggests that
there would be worthwhile gains in producing a processor that could offer two
different power-performance tradeoffs. The first configuration would maximise
the exploitation of ILP and consume the most power. The second would on
average perform less well, but would consume less power.
(i) Describe how the microarchitecture of a single processor could be modified
in order to offer the ability to switch between the two configurations
described at run-time. [9 marks]
(ii) How might one determine when to switch from one configuration to the
other in order to reduce overall power consumption while minimising the
impact on the user experience? [6 marks]