+ All Categories
Home > Documents > (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar...

(1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar...

Date post: 02-Jan-2016
Category:
Upload: julianna-curtis
View: 242 times
Download: 2 times
Share this document with a friend
9
(1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications by John Copeland, 2014) School of Electrical and Computer Engineering Georgia Institute of Technology
Transcript
Page 1: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(1)

ECE 3056: Architecture, Concurrency and Energy in

Computation

Lecture Notes by MKP and Sudhakar Yalamanchili

Sudhakar Yalamanchili(Some small modifications by John Copeland, 2014)

School of Electrical and Computer EngineeringGeorgia Institute of Technology

Page 2: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(2)

Course Objectives

• Basic concepts of microprocessor architecture Instruction Set Architecture (built on ECE 2020 and 2035

concepts) The Core: ALU, datapath, & control implementation Multiple Cores: synchronization Memory hierarchy: cache, virtual memory, disk Input/output systems

Basic Concurrency Supporto Multicore processorso Models of parallel computation

Sources of Energy and Power dissipationo Microarchitecture level models of energy and power

Image from http://blog.mytechhelp.com/laptop-repair/the-ivy-bridge/

Page 3: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(3)

Course Objectives

• Understand how computing systems work Fundamental architectural concepts and challenges Understanding performance: time, space, and energy

• Acquire knowledge to optimize systems As a designer As a user As a researcher

• Assignments Build simple processor models and apply several hardware

optimizations Build and simulate cache models Basic support for parallelism Evaluate energy consumption

Page 4: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(4)

Where Does the Computing Go?

From nytimes.com

Computer Engineering Inside

eecs.berkely.edu

Wireless sensor nodeGoogle Data Center

AMD Trinity APU

Page 5: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(5)

What You Will Learn

• How programs are translated into the machine language And how the hardware executes them

• The hardware/software interface

• What determines program performance And how it can be improved

• How hardware designers improve performance

• What is parallel processing?

• Where does the energy go?

Page 6: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(6)

Course Information

• Web page: http://users.ece.gatech.edu/~copeland/jac/3056/

And T-Square Will be constantly updated, so check it regularly

• Prerequisite: ECE2031, ECE 2035 or 3035

• Textbook Patterson and Hennessey, Computer Organization

& Design: The Hardware/Software Interface (4th edition, revised printing), Morgan Kaufmann, 2012. ISBN 978-0-12-374750-1

Page 7: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(7)

Grading Policy

• Homework Assignments: 30% Individual work, no collaboration No late assignments will be accepted You must make a passing grade on the assignments to pass

the course

• Exams 2 in-class exams: 40% (20% each, dates TBD) Final: 30%: December 13th, 2013, 2:50 pm – 5:40 pm

• Teaching Assistant: TBD

Page 8: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(8)

Where Are We?

• High-level language Level of abstraction closer

to problem domain Provides for productivity

and portability

• Assembly language Textual representation of

instructions

• Hardware representation Binary digits (bits) Encoded instructions and

data

ECE 2035

ECE 3056

Page 9: (1) ECE 3056: Architecture, Concurrency and Energy in Computation Lecture Notes by MKP and Sudhakar Yalamanchili Sudhakar Yalamanchili (Some small modifications.

(9)

Course Layout· Instruction set architecture· ALU Implementation· Basic Datapath· Pipelined Datapath· Chapter 2, 3, 4, Appendix B, C, D

· Cache design· Memory Hierarchy· Virtual Memory System· Chapter 5

· Storage Technologies· I/O Systems· Chapter 6

· Parallelism & Concurrency· Data, Thread, and Instruction Level Parallelism· Chapter 7, Appendix A

Intel Ivy Bridge


Recommended