Answer:
the floating-point instructions should be run 1.5 times faster in order to execute entire program 1.25 times faster
Explanation:
Given that:
a program consists of integer and floating-point instructions. 60% of the total execution time is spent on floating point instructions and the remaining 40% is on integer instructions.
Let the integer be the total execution time = V
The floating-point instructions = 60%
The integer instruction = 40%
The time spent on the floating-point instruction = 60/100 × V
= 0.6 V
The time spent on t he integer instruction = 40/100 × V
= 0.4 V
However; How much faster should we run the floating-point instructions to execute entire program 1.25 times faster
If we are to execute the entire program 1,25 times faster;
The new execution time = V/1.25
Assuming the new time spent on floating-point instruction = W
∵
W + 0.4 V = V/1.25
W = V/1.25 - 0.4 V
W = (V - 0.5V)/1.25
W = 0.4V
the new time spent on floating-point instruction = W = 0.4 V
The speed required to make the floating -point instruction to run faster = 0.6V/.4 V
= 1.5
Hence, the floating-point instructions should be run 1.5 times faster in order to execute entire program 1.25 times faster
