Performance Analysis [10 points (a)[5points] Consider a 5-stage MIPS pipeline where both branch target address is calculated and branching decision is resolved in ID stage. Assume that conditional branch represents 15 % of your code and branch is taken 60 % of time. The used dynamic branch prediction has 95 % accuracy. Calculate the additional stall cycles due to branch instructions (CPIstall branch) Note: No branch target buffer is used. Branch stage. prediction happens in IF (b) [5points] Assume the design in (a). Now we have a branch target buffer (BTB) that is accessed in IF stage. Calculate the additional stall cycles due to branch instructions stall branch Performance Analysis [10 points (a)[5points] Consider a 5-stage MIPS pipeline where both branch target address is calculated and branching decision is resolved in ID stage. Assume that conditional branch represents 15 % of your code and branch is taken 60 % of time. The used dynamic branch prediction has 95 % accuracy. Calculate the additional stall cycles due to branch instructions (CPIstall branch) Note: No branch target buffer is used. Branch stage. prediction happens in IF (b) [5points] Assume the design in (a). Now we have a branch target buffer (BTB) that is accessed in IF stage. Calculate the additional stall cycles due to branch instructions stall branch