loop:

lw $t$00($$s$1)$

add $t$0$ $t$1$ $t$2$

What is the performance improvement comparing single$-$cycle datapath and pipelined MIPS executions?

Assumptions:

$1.$ loop executes $32$ times.

$2.$ Each stage takes $46$ ns$.$