EE141-Fall 2012 Digital Integrated Circuits Instructor: Elad Alon TuTh 11-12:30pm 247 Cory 1 What is this class all about? Introduction to digital integrated circuit design engineering Will describe models and key concepts needed to be a good digital IC designer Models allow us to reason about circuit behavior Allow analysis and optimization of the circuit s performance, power, cost, etc. Understanding circuit behavior is key to making sure it will actually work Teach you how to make sure your circuit works Do you want your transistor to be the one that screws up a 1 billion transistor chip? 2 1
Detailed Topics CMOS devices and manufacturing technology CMOS gates Memories Propagation delay, noise margins, power Combinational and sequential circuits Interconnect Timing and clocking Arithmetic building blocks Design methodologies 3 What will you learn? Understanding, designing, and optimizing digital circuits for various quality metrics: Performance (speed) Power dissipation Cost Reliability 4 2
Practical Information Instructor Prof. Elad Alon 519 Cory Hall, 642-0237, elad@eecs Office hours: Tu. 3:30-4:30pm, Thurs. 2:30-3:30pm TAs: Alberto Puggelli, puggelli@eecs (OH: Wed. 4-5pm) Bonjern Yang, byang@eecs (OH: Wed. 3-4pm) Web page: http://bwrc.eecs.berkeley.edu/classes/icdesign/ee141_f12/ 5 Discussions and Labs Discussion sessions F 2-3pm (Bonjern) M 5-6pm (Alberto) Same material in all sessions! Labs (125 Cory) M 3-6pm (Bonjern) Tu 3-6pm (Alberto) Machines to left of double doors, with larger monitors Please choose one lab session and stick with it! 6 3
Your EECS141 Week 8 9 10 11 12 1 2 3 4 5 6 M Lab (Bonjern) 125 Cory DISC* (Alberto) 299 Cory T Lec (Elad) 277 Cory OH (Elad) 519 Cory Lab (Alberto) 125 Cory W OH OH (Bonjern) (Alberto) TBD Cory TBD Cory R Lec (Elad) 277 Cory OH (Elad) 519 Cory Problem Sets Due F DISC* (Bonjern) 299 Cory * Discussion sections will cover identical material 7 Class Organization 9 Assignments One design project (with a few phases) Labs: 5 software 2 midterms, 1 final Midterm 1: Thurs., October 4, evening (TBD) Midterm 2: Thurs., November 1, evening (TBD) Final: Wed., December 12, 8-11am 8 4
Some Important Announcements Please use piazza for asking questions (more later) Can work together on homework But you must turn in your own solution Lab reports due 1 week after the lab session Lab rules: http://california.eecs.berkeley.edu/iesg/labs/labinfo/labrules.asp Project is done in pairs No late assignments Solutions available shortly after due date/time Don t even think about cheating! 9 Grading Policy Homeworks: 12% Labs: 8% Projects: 20% Midterms: 30% Final: 30% 10 5
Class Material Textbook: Digital Integrated Circuits A Design Perspective, 2 nd ed, by J. Rabaey, A. Chandrakasan, B. Nikolic Class notes: Web page Lab Reader: Web page Check web page for the availability of tools 11 The Web Site http://bwrc.eecs.berkeley.edu/icdesign/eecs141_f10 Class and lecture notes Assignments and solutions Lab and project information Exams Many other goodies Print only what you need: Save a tree! 12 6
The Web Site #2 All announcements made at: https://piazza.com/berkeley/fall2012/ee141 Be sure to enroll! Piazza will also be the main forum for posting and answering questions Please post your questions there to minimize response time Also, check that your question hasn t been answered already 13 Software Cadence Widely used in industry Online tutorials and documentation HSPICE for simulation 14 7
Getting Started Assignment 1: Getting SPICE to work see web-page Due next Thursday, August 30, 5pm NO discussion sessions or labs this week. First discussion sessions in Week 2 First software lab in Week 3 15 Introduction Digital Integrated Circuit Design: The Past, The Present and The Future What made Digital IC design what it is today Why is designing digital ICs different today than it was before? Will it change in the future? 16 8
The First Computer The Babbage Difference Engine 25,000 parts cost: 17,470 17 ENIAC - The First Electronic Computer (1946) 18 9
The Transistor Revolution First transistor Bell Labs, 1948 19 The First Integrated Circuits Bipolar logic 1960 s ECL 3-input Gate Motorola 1966 20 10
Intel 4004 Microprocessor Intel, 1971. 2,300 transistors (12mm 2 ) 740 KHz operation (10 m PMOS technology) 21 Intel Pentium 4 Microprocessor Intel, 2005. 125,000,000 transistors (112mm 2 ) 3.8 GHz operation (90nm CMOS technology) 22 11
Intel Xeon (E7-8800) Intel, 2011. 2,600,000,000 transistors (513mm 2 ) 2.4 GHz operation (32nm CMOS technology) 23 Still Happening - Intel Test-Chip (2009) 22nm 364 MByte SRAM >2.91 billion transistors 24 12
Moore s Law In 1965, Gordon Moore noted that the number of transistors on a chip doubled every 18 to 24 months. He made a prediction that semiconductor technology will double its effectiveness every 18 months 25 Moore s Law LOG 2 OF THE NUMBER OF COMPONENTS PER INTEGRATED FUNCTION 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 Electronics, April 19, 1965. 26 13
Evolution in Complexity 27 Transistor Counts Doubles every 2 years 28 14
Frequency Frequency [MHz] 10000 1000 100 10 1 0.1 8008 4004 8080 Frequency Trends in Intel's Microprocessors 8086 8088 80286 486DX 386DX Pentium II Pentium Pro Pentium 486DX4 Pentium III Pentium MMX Pentium 4 Has been doubling every 2 years, but is now slowing down 1970 1975 1980 1985 1990 1995 2000 2005 29 Core2 Itanium II Itanium Power Dissipation Prediction (2000) Power (Watts) 100000 10000 1000 100 10 1 0.1 8085 8086286 386 486 4004 80088080 Pentium proc 18KW 5KW 1.5KW 500W 1971 1974 1978 1985 1992 2000 2004 2008 Year Did this really happen? Courtesy, Intel 30 15
Power [W] 1000 Power Dissipation Data 100 10 1 0.1 8008 4004 8080 8086 Power Trends in Intel's Microprocessors 8088 80286 Has been > doubling every 2 years 486DX 386DX Pentium Pro Pentium Pentium III Itanium II Itanium Pentium II Has to stay ~constant 1970 1975 1980 1985 1990 1995 2000 2005 31 Pentium 4 Core 2 Cause: Power Density Power Density (W/cm2) 10000 1000 100 10 1 4004 8008 8080 Sun s Surface Rocket Nozzle Nuclear Reactor 8086 Hot Plate 8085 286 386 486 P6 Pentium proc 1970 1980 1990 2000 2010 Year Power density too high for cost-effective cooling S. Borkar 32 16
Not enough cooling *Pictures from http://www.tomshardware.com/2001/09/17/hot_spot/ 33 Not Only Microprocessors Cell Phone Small Signal RF Power RF Units Digital Cellular Market (Phones Shipped) 1996 1997 1998 1999 2000 48M 86M 162M 260M 435M Power Management Analog Baseband Digital Baseband (DSP + MCU) (data from Texas Instruments) 34 17
Challenges in Digital Design DSM Microscopic Problems Ultra-high speed design Interconnect Noise, Crosstalk Reliability, Manufacturability Power Dissipation Clock distribution. Everything Looks a Little Different? 1/DSM Macroscopic Issues Complexity Time-to-Market Millions of Gates High-Level Abstractions Reuse & IP: Portability Predictability etc. and There s a Lot of Them! 35 Productivity Trends 10,000,000 10,000 1,000,000 1,000 100,000 100 10,000 10 1,0001 100 0.1 0.01 10 Logic Tr./Chip Tr./Staff Month. x x x x x x x x 58%/Yr. compounded Complexity growth rate 21%/Yr. compound Productivity growth rate 100,000,000 10,000,000 1,000,000 100,000 10,000 1,0001 100 0.1 0.001 1 10 0.01 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009 Complexity Logic Transistor per Chip (M) Productivity (K) Trans./Staff - Mo. Source: Sematech Complexity outpaces design productivity Courtesy, ITRS Roadmap 36 18
Why Scaling? Technology shrinks by 0.7/generation With every generation can integrate 2x more functions per chip; chip cost does not increase significantly Cost of a function decreases by 2x But How to design chips with more and more functions? Design engineering population does not double every two years Hence, a need for more efficient design methods Exploit different levels of abstraction 37 Design Abstraction Levels SYSTEM + MODULE GATE CIRCUIT S n+ G DEVICE D n+ 38 19
Next Lecture Introduce basics of integrated circuit manufacturing and cost 39 20