Download presentation
Presentation is loading. Please wait.
Published byLouisa Baker Modified over 8 years ago
1
MS108 Computer System I Lecture 5 Pipeline Prof. Xiaoyao Liang 2015/3/27 1
2
Pipeline MIPS 2
3
3Example In this example: –Sequential execution takes 4 * 90min = 6 hours –Pipelined execution takes 30+4*40+20 = 3.5 hours Bandwidth = loads/hour BW = 4/6 l/h w/o pipelining BW = 4/3.5 l/h w pipelining BW <= 1.5 l/h w pipelining, more total loads Pipelining helps bandwidth but not latency (90 min) Potential speedup = Number pipe stages ABCD 6 PM 789 TaskOrderTaskOrder Time 3040 20 Dave Patterson’s Laundry example: 4 people doing laundry wash (30 min) + dry (40 min) + fold (20 min) = 90 min Latency
4
4 Pipeline Datapath
5
5 Ideal Pipeline
6
Divide the Datapath 6
7
Timing Assumption 7
8
5-Stage Pipeline 8
9
Resource Utilization 9
10
ALU Instructions 10
11
11 IR and Control
12
12 Control Generation
13
13Hazards
14
14 Data Hazard
15
15Solutions
16
Feedback Killing 16
17
Stall 17
18
Stall and Pipeline Bubble 18
19
19 Generate Stall
20
20 Generate Stall
21
21 Potential Stall Instructions
22
22 Generate Stall
23
23 Load/Store Hazard
24
24 Load/Store Hazard
25
25Jump
26
26 Pipelined Jump
27
27 Pipelined Jump
28
28Branch
29
29Branch
30
30Branch
31
31 New Stall
32
32 Control Signals
33
33 Branch Pipeline
34
34 Reducing Branch Penalty
35
35 Branch Delay Slot
36
36Bypassing
37
37 Adding Bypassing
38
38 Fully Bypassing Datapath
39
39 Limitation of Bypassing
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.