Which statement correctly describes the partitioning within a $128-$bit adder used for subword parallelism?

The adder is used to perform a single operation at a time on $128-$bit data.

The adder is partitioned to store multiple data types simultaneously.

The adder is solely dedicated to floating point operations.

The adder is partitioned to perform parallel operations on multiple smaller data sizes.