Consider the operation of a machine with the data path in the figure below Suppose that loading the ALU input registers takes 10 ns, running the ALU takes 10 ns, and storing the result back in the register scratchpad takes 10 ns. What is the maximum number of MIPS this machine is capable of in the absence of pipelining? What is the speedup benefit possible from pipelining