Introduction to Computer Systems Engineering

1
Introduction to Computer Systems Engineering

• Stamatis Vassiliadis / Georgi N. Gaydadjiev/
  Georgi Kuzmanov
• Computer Engineering Laboratory
• EEMCS
• Delft University of Technology

2
Introduction

• Code ET4246
• Docents Stamatis Vassiliadis / Georgi N.
  Gaydadjiev
• E-mails stamatis_at_ce.et.tudelft.nl /
  georgi_at_ce.et.tudelft.nl /
• G.Kuzmanov_at_ewi.tudelft.nl
• URLs ce.et.tudelft.nl/stamatis /
  ce.et.tudelft.nl/georgi /
• ce.et.tudelft.nl/george
• Exam pass/fail, multiple-choice
• Office 15.250 / 15.320 / 15.280
• office hours 1145-1245 after class or after
  appointment
• Things youll be learning
• what the Computer Engineering field is, a basic
  foundation
• what are the basic components and the main ideas
  used
• prepare you for the courses to follow
• Book Lecture Notes ET4246
• Additional reading see Recommended Literature
  list in the lecture notes

3
Typical Lecture Format

• 20 minutes Lecture
• 5 minutes Questions
• 15 minutes Lecture
• 5 minutes Questions
• If you have no questions then 45 minutes
  lecture!!!!

• OUR GOALS
• Show you what the basic concepts are
• Show you how those concepts are applied to
  computers
• NOT to talk at you
• so...
• ask questions
• come to office hours
• find us in the Computer Engineering Lab (15th
  floor)
• ...

4
ET4 246 in context

• Prerequisites
• BSc in Computer Science or
• BSc in Electrical Engineering
• This course will introduce you to the main topics
• System software
• Hardware
• Design Tools
• Follow-on courses see your MSc program

5
Where we are headed?

• Software (Chapter 2) Compilers and Operating
  Systems
• Hardware (Chapter 3) From Architecture down to
  logic design
• VLSI Design Tools and methodologies (Chapter 4)
  how the physical machines are realized
• Key to passing the exam reading the lecture
  notes!

6
Why studying Computer Engineering?

• To become a computer designer
• Computer Engineers designed your computer and
  more
• To understand how computing systems operate and
• Write better system Software (Compilers, O/S
  etc.)
• Design complex computer based systems
• Create new applications for computing systems
• Because it is intellectually fascinating!
• Computer is the most complex man-made device

7
What Computer Engineers Do?

• Apply
• computing, mathematics and engineering
  theories and principles to the
• design of
• computer hardware, software, networks and
  computerized equipment
• to solve technical problems
• in diverse application domains.

8
Computer- as we know it now
Did computers always looked like this?
display
network connection
keyboard
Will computers always look like that?
9
The first known computing device
Built around 87 B.C.
Discovered in 1901
Used for astronomical calculations
Sophisticated 20 gears assembly
Knowledge completely lost
10
History

• 76 B.C. - Antikythera mechanism
• 1642 Blaise Pascal - calculating machine
• 1800s Babbadge analytical engine
• 1930s 1940s electromechanical computers
• 1937 J. Atanasoff / C. Berry ABC (first
  electronic computer)
• 1943/46 J. Mauchly et al - ENIAC (18,000 tubes)
• 1945 J von Neumann von Neumann architecture
• 1948, 1958 transistor IC invention
• 1960s transistor based computers
• 1971 Intel 4004 the first GP microprocessor
• The modern computer era started

11
History (software)

• 1800s Ada Byron concepts of branches and loops
• 1945 Konrad Zuse Plankalkul (first algorithmic
  language)
• 1949 J. Mauchly Short Code (first HLL)
• 1951 G. Hopper the first compiler
• 1957 J. Backus FORTRAN
• 1960s IBM virtual machine concept
• 1968 N. Wirth Pascal
• 1970s ATT - UNIX
• 1972 D. Ritchie et al C
• 1987 A. Tanenbaum MINIX (open source clone of
  UNIX)
• 1991 L. Torvalds Linux

12
Classes of Computing Devices

• Supercomputer 5-20 million
• Mainframe 0.5-4 million
• Server 10-200 thousand
• PC/Workstation 1-10 thousand
• Game console 300-1000
• Embedded device 1-100
• Future disposable 1-100 cents

13
Computers Turing view
14
Solving Computational Problems
Problem
Solutions
15
Abstraction

• Difference between interface and implementation
• Interface WHAT something does
• Implementation HOW it does so

16
Abstraction simple example

• 21 Multiplexer (MUX)
• Interface
• Implementations
• Gates (fast or slow), pass transistors

X
Y
Mux
S
F
17
Abstraction and Complexity

• Abstraction helps to manage complexity
• Complex interfaces
• Specify what to do
• Hide details of how

Design Tools and Methodologies

• Main goal remove magic

18
Why is Abstraction important?

• Complex interfaces implemented by layers below
• Abstraction hides detail
• Hundreds of engineers build one product
• Shorter development times
• Complexity unmanageable otherwise

19
Computer Engineering

• Exercise in engineering tradeoff analysis
• Find the fastest/cheapest/power-efficient/etc.
  solution
• Optimization problem with 1000s of variables
• All the variables are changing continuously
• At non-uniform rates
• With inflection points
• Only one guarantee Todays right answer will be
  wrong tomorrow!
• Two high-level effects
• Technology Push
• Application Pull

20
Technology Push

• What do these two intervals have in common?
• 1700s-1999 (300 years)
• 2000-2001 (2 years)

• Answer Equal progress in processor speed!

• The power of exponential growth!
• Driven by Moores Law
• Devices per chip doubles every 18-24 months
• Computer engineers work to turn the additional
  resources into speed/power savings/functionality!

21
Moore's Law

• Moores law is good for many years!

22
Technology Push

• Technology advances at varying rates
• E.g. DRAM capacity increases at 60 / year
• But DRAM speed only improves 10 / year
• Creates gap with processor frequency!
• Inflection points
• Crossover causes rapid change
• E.g. enough devices for single-chip processor
• Imminent system on a chip (SoC) and chip
  multiprocessors (CMP)
• Imminent clock signal cannot reach entire chip

23
Application Pull

• Corollary to Moores Law
• Cost halves every two years
• In a decade you can buy a computer for less than
  its sales tax today. Jim Gray
• Computing devices cost-effective for
• Enterprise computing banking, stock exchange
• Departmental computing computer-aided design
• Personal computer spreadsheets, email, web
• Embedded systems look around you

24
Application Pull

• What about the future?
• E.g. weather forecasting computational demand
• Applications that are not cost-effective today
• Virtual reality
• Teleconferencing
• New Web technologies
• Global Wireless Networks
• Proactive (beyond interactive) devices
• This is your job!

25
Performance vs. Design Time

• Time to market is very important (almost
  critical)
• E.g. , new design development may take 3 years
• It will be 3 times faster than other
• But if technology improves 50 / year
• In 3 years 1.53 3.375
• So the new design is worse than existing ones!
• (unless it also employs new technology)

26
Bottom Line

• Designers must know BOTH software and hardware
• Their contribution to the layers of abstraction
• IC costs / performance aspects
• Compilers and Operating Systems
• Design Tools and new Design Methods

27
Current trends

• Parallelism in microprocessors
• Multithreaded execution
• SIMD parallelism
• Explicit instruction-level parallelism
• Reconfigurable computing
• Low-power portable computing
• Reducing the energy consumed by microprocessors
• Computing in laptops, handheld devices, watches
  (e.g. IBMs Linux watch project!), sensors
  (ambient computing),
• Internetworking and ubiquity
• Services available over wired or wireless
  networks

28
Final Notes

• Next week - Compilers / Operating Systems
• Lecture notes will be available