L15 - 18-447 Lecture 15: A Whirlwind Tour of Modern...

Info iconThis preview shows pages 1–5. Sign up to view the full content.

View Full Document Right Arrow Icon
CMU 18-447 S’08 L15-1 © 2008 J. C. Hoe 18-447 Lecture 15: A Whirlwind Tour of Modern Microarchitectures James C. Hoe Dept of ECE, CMU March 19, 2008 Announcements: Project, project, project, Midterm, midterm, midterm This lecture won’t be covered on the midterm or the final Handouts: The Microarchitecture of Superscalar Processors, Smith and Sohi, Proceedings of IEEE, 12/1995. (on Blackboard) Practice Midterm CMU 18-447 S’08 L15-2 © 2008 J. C. Hoe Performance Factors T wall-clock = T cyc × CPI × No.Instructions cycles-per-instruction max. combinational delay ISA and compilers
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
CMU 18-447 S’08 L15-3 © 2008 J. C. Hoe Going after IPC Scalar Pipeline (baseline) Operation Latency = 1 Peak IPC = 1 Instruction-Level Parallelism = 1 IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB instruction stream base cyc 0 1 2 3 4 5 6 7 8 9 10 CMU 18-447 S’08 L15-4 © 2008 J. C. Hoe Superpipelined Machine Superpipelined Execution OL = 1 baseline cycle (M minor cycles) Peak IPC = M per baseline cycle (1 per minor cycle) ILP = M major cycle = M minor cycles minor cycle base cyc 0 1 2 3 4 5 6 7 8 9 10 IF ID EX MEM WB IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF
Background image of page 2
CMU 18-447 S’08 L15-5 © 2008 J. C. Hoe Superscalar Machines Superscalar (Pipelined) Execution OL = 1 baseline cycles Peak IPC = N per baseline cycle ILP = N IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB instruction stream base cyc 0 1 2 3 4 5 6 7 8 9 10 IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB IF ID EX MEM WB CMU 18-447 S’08 L15-6 © 2008 J. C. Hoe Superscalar Datapath I-cache Reg File Read PC D-cache ALU ALU Reg File Write 2 X fetch bandwidth 2 X read ports 2 X Logic Can’t always double resources 2 X write ports Pipe A Pipe B ?
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
CMU 18-447 S’08 L15-7 © 2008 J. C. Hoe Superscalar and Superpipelined Achieving peak performance on each architecture depends on finding N or M independent instructions per cycle Superscalar Parallelism Operation Latency: 1 Issuing Rate: N Superscalar Degree: N Superpipeline Parallelism Operation Latency: 1 Issuing Rate: M Superpipelined Degree: M instruction stream IF ID MEM WB IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF IF ID MEM WB IF ID MEM WB IF ID MEM WB IF ID MEM WB IF ID MEM WB IF ID MEM WB vs.
Background image of page 4
Image of page 5
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 04/07/2008 for the course ECE 18447 taught by Professor Hoe during the Spring '08 term at Carnegie Mellon.

Page1 / 15

L15 - 18-447 Lecture 15: A Whirlwind Tour of Modern...

This preview shows document pages 1 - 5. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online