instruction and data memories data formats

Question

I am confused on the wording of the following statement for a Computer Architecture problem:

The data path must have separate instruction and data memories because the formats of data and instructions are different in MIPS and hence different memories are used.

I know that each data path element can only do one function at a time, thus, I need separate instruction and data memories, but is this the same as saying the previous statement?

Answer 1

It seems somewhat off-base to me.

"The data path must have separate instruction and data memories because the formats of data and instructions are different in MIPS and hence different memories are used" says to me that the reason you can't access both the data and instruction memory at the same time is because they store different information, which is irrelevant.

As you said, the primary reason why memory is divided between instruction and data in a single-cycle machine is that memory can only be accessed once per cycle.

Answer 2

Each component (memory, ALU, etc.) can be used only once in a single clock cycle. Therfore, the real reason for assuming separate instruction and data memories is that we cannot access memory components multiple times during each single clock cycle.