0 votes 0 votes https://gateoverflow.in/3437/gate2007-it-6-isro2011-25 Why are they not calculating average CPI in pipelined processor and then calculating speedup ??? CO and Architecture speedup cpi + – daksirp asked Sep 2, 2018 daksirp 458 views answer comment Share Follow See all 2 Comments See all 2 2 Comments reply MiNiPanda commented Sep 2, 2018 reply Follow Share speedup= $\frac{np}{p}$ np=time taken by n instructions in a non-piplelined system p=time taken by n instructions in a piplelined system Can you show your approach? 0 votes 0 votes daksirp commented Sep 2, 2018 reply Follow Share in my context, speedup = Twp / Tp {Twp : time without pipelining, Tp : time with pipelining} Twp = n * 12 clocks, Tp = there are multiple stages with execution times 3, 2, 5, 4, 6, 2 clocks. therefore we need to calculate average CPI. avg CPI = { [m * 1] + [1 * (n - 1)] } / n , [m is number of stages], [n is number of instructions ] Tp = Avg CPI * T(for 1 clock cycle, which is max time ==> 6 clocks) i did a mistake of assuming 'n' no of instructions to be any number since we are taking average, it can be any number. i didnt read the que nicely, there it is given 'n' as very large.... so that formula for Tp becomes [n * (max time for a stage)]. Now i understood. thnx. 1 votes 1 votes Please log in or register to add a comment.