Amdahl’s Law CPS 5401 Fall 2013 Shirley Moore

Slides:



Advertisements
Similar presentations
You WANT me to make a paper airplane??? A lesson in calculating the speed of an object.
Advertisements

Performance Evaluation of Architectures Vittorio Zaccaria.
TU/e Processor Design 5Z032 1 Processor Design 5Z032 The role of Performance Henk Corporaal Eindhoven University of Technology 2009.
CS 6290 Evaluation & Metrics. Performance Two common measures –Latency (how long to do X) Also called response time and execution time –Throughput (how.
Computer Organization and Architecture 18 th March, 2008.
Performance D. A. Patterson and J. L. Hennessey, Computer Organization & Design: The Hardware Software Interface, Morgan Kauffman, second edition 1998.
Copyright © 1998 Wanda Kunkle Computer Organization 1 Chapter 2.5 Comparing and Summarizing Performance.
Computer Performance Evaluation: Cycles Per Instruction (CPI)
Recap.
CIS429.S00: Lec2- 1 Performance Overview Execution time is the best measure of performance: simple, intuitive, straightforward. Two important quantitative.
Computer Architecture Lecture 2 Instruction Set Principles.
Chapter 4 Assessing and Understanding Performance
Chapter 4 Performance. The Role of Performance Example of Performance Measure.
Fall 2001CS 4471 Chapter 2: Performance CS 447 Jason Bakos.
Amdahl's Law.
1 Chapter 4. 2 Measure, Report, and Summarize Make intelligent choices See through the marketing hype Key to understanding underlying organizational motivation.
GCSE Maths (Higher Tier) Inverse Proportion. Direct proportion what does it mean? £ Percentage Both.
1 ECE3055 Computer Architecture and Operating Systems Lecture 2 Performance Prof. Hsien-Hsin Sean Lee School of Electrical and Computer Engineering Georgia.
1 Measuring Performance Chris Clack B261 Systems Architecture.
CMSC 611: Advanced Computer Architecture Performance Some material adapted from Mohamed Younis, UMBC CMSC 611 Spr 2003 course slides Some material adapted.
1 Computer Performance: Metrics, Measurement, & Evaluation.
Gary MarsdenSlide 1University of Cape Town Computer Architecture – Introduction Andrew Hutchinson & Gary Marsden (me) ( ) 2005.
Performance Evaluation of Parallel Processing. Why Performance?
Copyright 1995 by Coherence LTD., all rights reserved (Revised: Oct 97 by Rafi Lohev, Oct 99 by Yair Wiseman, Sep 04 Oren Kapah) IBM י ב מ 7-1 Measuring.
CDA 3101 Fall 2013 Introduction to Computer Organization Computer Performance 28 August 2013.
1 CS/EE 362 Hardware Fundamentals Lecture 9 (Chapter 2: Hennessy and Patterson) Winter Quarter 1998 Chris Myers.
Compiled by Maria Ramila Jimenez
Introduction The average running speed of a human being is 5–8 mph. Imagine you set out on an 8-mile run that takes you no longer than 1 hour. You run.
Computer Architecture
Lecture 8: 9/19/2002CS170 Fall CS170 Computer Organization and Architecture I Ayman Abdel-Hamid Department of Computer Science Old Dominion University.
Performance – Last Lecture Bottom line performance measure is time Performance A = 1/Execution Time A Comparing Performance N = Performance A / Performance.
CEN 316 Computer Organization and Design Assessing and Understanding Performance Mansour AL Zuair.
CPU Performance using Different Parameters CS 250: Andrei D. Coronel, MS,CEH,PhD Cand.
Performance Enhancement. Performance Enhancement Calculations: Amdahl's Law The performance enhancement possible due to a given design improvement is.
 You can use weighted averages to solve uniform motion problems when the objects you are considering are moving at constant rates or speeds.
1  1998 Morgan Kaufmann Publishers How to measure, report, and summarize performance (suorituskyky, tehokkuus)? What factors determine the performance.
Performance Performance
TEST 1 – Tuesday March 3 Lectures 1 - 8, Ch 1,2 HW Due Feb 24 –1.4.1 p.60 –1.4.4 p.60 –1.4.6 p.60 –1.5.2 p –1.5.4 p.61 –1.5.5 p.61.
September 10 Performance Read 3.1 through 3.4 for Wednesday Only 3 classes before 1 st Exam!
Performance – Last Lecture Bottom line performance measure is time Performance A = 1/Execution Time A Comparing Performance N = Performance A / Performance.
Lecture 5: 9/10/2002CS170 Fall CS170 Computer Organization and Architecture I Ayman Abdel-Hamid Department of Computer Science Old Dominion University.
L12 – Performance 1 Comp 411 Computer Performance He said, to speed things up we need to squeeze the clock Study
Computer Engineering Rabie A. Ramadan Lecture 2. Table of Contents 2 Architecture Development and Styles Performance Measures Amdahl’s Law.
Proportionality SPH4U. Introduction In physics, we are often interested in how one variable affects another.
CMSC 611: Advanced Computer Architecture Performance & Benchmarks Some material adapted from Mohamed Younis, UMBC CMSC 611 Spr 2003 course slides Some.
Performance Computer Organization II 1 Computer Science Dept Va Tech January 2009 © McQuain & Ribbens Defining Performance Which airplane has.
Performance. Moore's Law Moore's Law Related Curves.
September 2 Performance Read 3.1 through 3.4 for Tuesday
ECE 4100/6100 Advanced Computer Architecture Lecture 1 Performance
Defining Performance Which airplane has the best performance?
Speed and Velocity.
Chapter 11, Section 2 Speed & Velocity
Prof. Hsien-Hsin Sean Lee
Introduction The average running speed of a human being is 5–8 mph. Imagine you set out on an 8-mile run that takes you no longer than 1 hour. You run.
Chapter 1 Fundamentals of Computer Design
CS2100 Computer Organisation
Parallel Computers.
Defining Performance Section /14/2018 9:52 PM.
Parallel Processing Sharing the load.
CMSC 611: Advanced Computer Architecture
EQ: How Can We Measure Speed?
Amdahl's law.
Proportionality SPH4U.
Performance Cycle time of a computer CPU speed speed = 1 / cycle time
CMSC 611: Advanced Computer Architecture
January 25 Did you get mail from Chun-Fa about assignment grades?
Parameters that affect it How to improve it and by how much
Performance Measurement and Analysis
Chapter 2: Performance CS 447 Jason Bakos Fall 2001 CS 447.
CS2100 Computer Organisation
Presentation transcript:

Amdahl’s Law CPS 5401 Fall 2013 Shirley Moore

Amdahl’s Law The performance improvement to be realized from using some faster mode of execution is limited by the fraction of time the faster mode can be used. “The performance improvement to be realized from improving some feature of a design is limited by the amount of time that feature is used and by the amount of improvement to the feature.” “Make the common case fast.”

Speedup What is speedup? Performance for entire task using enhancement Speedup = ______________________________________ Performance for entire task without enhancement OR Execution Time without enhancement _______________________________ Execution Time with enhancement Realizing that Performance is inversely related to execution time: e.g., to compare performance of two machines:

Amdahl’s Law provides a way to calculate the speedup to be realized from some design enhancement which depends on two factors: Fraction of computation time in original machine that will be affected by the enhancement. The amount of said enhancement, ie. “how much improvement did we make to the feature under consideration?” One might find that the desired speedup cannot be obtained by improving a single feature. Consider the following common riddle: You are traveling from Point A to Point C going through Point B. The distance from A-> B is 1 mile and the distance from B -> C is one mile. You travel from A to B at 30 mph, how fast must you go in traveling from Point B to Point C in order to average 60 mph? Amdahl’s Law and Speedup

A machine takes 100 sec to execute a given program 40 seconds for fp 60 seconds for everything else Question: Can I double the speed by enhancing fp? Suppose that I improve fp by a factor of 10, what is the speedup that can be realized? Example 1

Example 2 Amdahl’s Law expresses the Law of Diminishing Returns Suppose I have a choice of two improvements. Fp square root = 20% or all fp 50% Proposition 1 -- improve square root by a factor of 10 Proposition 2 --improve all fp by a factor of 2 Which would you choose? Prop 1. Speedup = 1 _______________ = /10 Prop 2. Speedup = 1 ________________ = /2 It is sometimes difficult to determine how much time is being used by particular components.