a$)$

$n=1$;

for $j=1$ to $100$

if $(n>10)$

I

for $i=1$ to $100$

end for

$n=1$;

else

I

for $i=1$ to $100$

$.1$

end for

$n=n+1$;

end if

end for

For the above code, identify which branch prediction scheme would result in fewer additions to the

CPI for the pipelined processor of Q$1($Branch condition is evaluated in EX satge$).$ Static branch

prediction, dynamic branch prediction with $1-$bit branch history table and with $2-$bit branch history

table. Justify you answer by giving examples from the code where performance is improved by using

a particular scheme over the other.