retagged by
3,115 views
2 votes
2 votes
Consider a 5 stage instruction pipeline having latencies (in ns) 1, 2, 3, 4 and 5 respectively.
Find average CPI of non-pipeline CPU when speed up achieved with respect to pipeline is 4
(assume ideal case for pipelining)?
(a) 1.23
(c) 1.66
(b) 1.33
(d) 1.73
retagged by

1 Answer

Best answer
9 votes
9 votes

Tnon-pipeline = sum of all stages delay = 15ns

Tpipeline  = max delay among all stages delay = 5ns

Given speedup = 4

Speedup = $\frac{\text{Execution of non pipeline}}{\text{Execution of pipeline}}$

Execution of non pipelined = CPInon-pipelined * Cycle time

Execution of non pipelined = CPInon-pipelined * 15ns

Now,

Execution of pipelined = CPIpipelined * cycle time 

In pipelined ideally CPIpipelined = 1

Execution of pipelined = 1 * 5ns

4 = (CPInon-pipelined *15)/(1*5)

CPInon-pipelined = 4/3

CPInon-pipelined = 1.33

Ans = 1.33

selected by

Related questions

692
views
1 answers
0 votes
kathan Mistry asked Aug 9, 2022
692 views
In a pipelined architecture Static branch prediction is used with branch taken assumption. Assume that 30% of the instructions executed for a program are branch ... Execution time(in nsec) if 80 % of the instructions are taken?
2.2k
views
2 answers
0 votes
Sourajit25 asked Nov 26, 2017
2,194 views
What is the definition of an ideal pipeline?On searching over the internet I found only speed up formulas but no where the difference between ideal and non-ideal was mentioned.
15.2k
views
1 answers
12 votes
PEKKA asked Oct 27, 2016
15,211 views
Collition Vector : 1011010 MAL for the above Collition Vector is _____Please also tell me how to calculate efficiency and throughtput
2.8k
views
2 answers
2 votes
biranchi asked Nov 17, 2016
2,766 views
Consider a non-pipelined processor design which has a cycle time of 10ns and average CPI of 1.4. If each pipeline stage adds extra 20ps due to register setup ... the time for 2 cycles (these occurrences are disjoint). What is the new CPI?