Here cycle time = 1 / (1 GHz) = 1 ns
Here CPInon branch = 1 which is default assumption also (CPI of an ideal pipeline is 1)..
Time taken by 109 instructions = 1.6 s
Hence time taken by 1 instruction = 1.6 ns
Hence number of cycles taken by 1 instruction = Time taken by 1 instruction / Cycle time = 1.6
Hence CPIeffective = 1.6
But we know CPIeffective = 1 + Number of Stalls / instruction
= 1 + (fraction of branch instruction * branch penalty)..
= 1 + x * branch penalty
Now
Branch penalty = Stage at which target address is found for the branch location - 1
Now the stage is given as 5th stage..Hence branch penalty = 4
Hence
CPIeffective = 1 + x * 4
==> 1.6 = 1 + 4x
==> 4x = 0.6
==> x = 0.15 = 15 %
Hence number of branch instructions = (15/100) * 109
= 15 * 107 branch instructions