Transcript The Course`s Goals

The Course’s Goals
To be interesting and fun.
An interested student learns more.
To answer questions that are probably
keeping you awake at night: What is a RISC
machine? How does the cache work? Who
was Von Neumann?
For you to write better programs.
By understanding how your program is
executed on a computer, will enable you to
program better.
Computer
Structure - Introduction
1/17
Not the Course’s Goals
To be b o r i n g. A bored student:
a. Fall asleep, that’s OK just don’t snore.
b. Talks, that’s not OK.
To answer questions that aren’t related to
computers: “Will the Lakers repeat?” has
nothing to due with the course.
For you to design computers.
You will need much more than this course but
it is a step in the right direction.
Computer
Structure - Introduction
2/17
 The main goal of the course is to help you understand how
the computer executes your programs.
 Thus we won’t describe the computer from a transistor level
but from the view point of your program.
 We will show how your program is translated into machine
language and how the processor executes the instructions.
 After you have taken this course and an advanced
architecture course, you will have a background in which you
can start to understand the design of a whole computer
system.
 If you can’t keep quiet during class please leave. You don’t
have to participate in class. I will remove students from class
if necessary.
General Information
Contact is by e-mail ([email protected]),
phone (6751168 (jct), 6585766 (hu), 6786784
(home)),or in person (look at my home page for
details).
Grade: the final grade is composed of a final
exam (70%) and three mini-exams (30%), out
of which the 2 best will be used.
Assignments are optional. They will be given
in the same format as the mini-exams.
Computer
Structure - Introduction
3/17
Goal: Describe a Computer
What is a computer?
Which of the following is a computer?
How do we define a computer?
Computer
Structure - Introduction
4/17
 Shown are (clockwise from right):
 Charles Babbage’s Difference Machine (1822) designed
for the purposes of computing the entries in navigational
and other tables (even received the first government
grant for computer research).
 Motorola’s Power PC 604e microprocessor.
 Standard PC motherboard.
 The Abacus (‫ )חשבונית‬which is generally considered to be
the first mechanical computer.
 Blaire Pascal’s Adding Machine (1642) with automatic
carries from one position to the next.
The 5 Classic Components
Data flows from the input devices into memory, from
the memory into the processor. The data is then
processed and written back to memory. It is then
stored or displayed
Computer
in the output
Processor
devices.
Memory
Devices
Computer
Control
Input
Datapath
Output
Structure - Introduction
5/17
 That is, any computer, no matter how primitive or advance,
can be divided into five parts:
 1. The input devices bring the data from the outside world
into the computer.
 2. These data are kept in the computer’s memory until ...
 3. The datapath request and process them.
 4. The operation of the datapath is controlled by the
computer’s controller.
 All the work done by the computer will NOT do us any good
unless we can get the data back to the outside world.
 5. Getting the data back to the outside world is the job of the
output devices.
 The five components can reduced to 3 basic components
(like in the army): Processor, Memory, and I/O.
The Processor
The “heart” (and brain) of the computer.
Schedules instruction execution (the Control)
and executes the instructions (the Datapath).
Branch
Control
Data
cache
Bus
Instruction
cache
Computer
Integer
datapath
Floatingpoint
datapath
Structure - Introduction
6/17
Performance Increase
The rate of performance improvement doubles every
1.6 years, or 1.54 per year. The following graph
shows the improvement of workstations over a
period of 10 years.
1200
DEC Alpha 21264/600
1100
1000
900
Performance
800
700
600
500
DEC Alpha 5/500
400
300
DEC Alpha 5/300
200
100
SUN-4/ MIPS
260
M/120
0
1987
Computer
1988
MIPS
M2000
1989
IBM
RS6000
1990
DEC Alpha 4/266
IBM POWER 100
DEC AXP/500
HP 9000/750
1991
1992
Year
1993
Structure - Introduction
1994
1995
1996
1997
7/17
 The processor reads instructions and data from memory and
“processes” (‫ )מעבד‬it.
 The processor shown is the Intel Pentium.
 All modern processors since the i486 have onchip caches
and floating point units.
 In 1965 Gordon Moore (co-founder of Intel) predicted that
the number of transistors on chip will double every 18
months. He was right. This is called “Moore’s Law”, it is one
of the reasons for the performance growth.
 But not the only reason. Performance doubles every 16
months due to improvements in design as well.
Memory and the Motherboard
The CPU, memory chips and other devices
sit in slots on the motherboard
Audio/
MIDI
Four
ISA
card
slots
Four
PCI
card
slots
Parallel/
serial
Processor
Four
SIMM
slots
Two IDE
connectors
Computer
Structure - Introduction
8/17
Memory Capacity Growth
Memory capacity has quadrupled (4 ‫) גדל פי‬
every 3 years. The following chart shows how
SIMMs have grown over a period of 20 years.
100,000
64M
16M
Kbit capacity
10,000
4M
1M
1000
256K
100
64K
16K
10
1976
1978
1980
1982
1984
1986
1988
1990
1992
1994
1996
Year of introduction
Computer
Structure - Introduction
9/17
 On the motherboard sit the processor, caches, memory
SIMMs (Single Inline Memory Modules), and the interfaces
to the I/O devices.
 They are connect by a series of Busses (‫)ערוצים‬, the ISA and
PCI bus are standards for most PCs.
 The memory growth rule is called the “DRAM growth rule”.
 Today PCs are equipped with 64-128 Mbytes (million bytes)
of memory. High end workstations can have even Gigabytes
(billions of bytes) of memory.
 The memory contains the programs and data the processor
works with.
I/O Devices
I/O devices can be divided into 3:
Input - keyboard, mouse, joystick, scanner,
CDROM, DVD, microphone, digital camera ...
Output - printer, display, barcode projector,
speakers ...
Input/Output - hard disks, floppy disks, magnetic
tapes, VR (Virtual reality) helmets and gloves …
Computer
Structure - Introduction
10/17
 Most devices are becoming both Input and Output.
 Touch screens are becoming very common. There are even
screens which you can activate with a beam of light.
 Mice and Joysticks are becoming more and more interactive.
 Most read only media such as CDs and DVDs are writable
as well.
 Hard disks, floppy disks, magnetic taps, Zip and Jaz disks,
optical disks, DVDs, CDs, and many more are called
storage devices as well.
 The I/O devices are the only way for us humans to interact
with the computer.
Modern Computers
Are these computers?
Computer
Structure - Introduction
11/17
 Shown are (clockwise from right):
 A “wearable computer”. The computer is in the persons
pocket. He interacts with the computer through a headset
which contains a screen, microphone, and earphones. In his
hand is a trackball mouse.
 The Cassiopeia is a handheld device (like the Palm Pilot)
which can runs a versions of Windows (Windows CE).
 Sony’s WebTV enables interactive TV viewing and web
surfing.
 The Cray XMP (1976) was a super cooled (the orange
towers are filled with liquid nitrogen) Super Computer.
ENIAC - The First Electronic Computer
Computer
Structure - Introduction
12/17
ENIAC was a general purpose computer used for
computing artillery tables.
It was U shaped, 25m long, 2.5m high and 1m wide
ENIAC used 18,000 vacuum tubes.
Programming was done by plugging cables and
setting switches. Data was entered by punched
cards.
Programming for typical calculations took from half
a hour to a day.
ENIAC was 2 magnitudes (‫ ) סדר גודל‬larger and 4
magnitudes slower than modern computers (1900
adds per sec).
Goal: Describe Computing
We think we know what a computer is.
What is computing?
Idea
Computer
Programmer
Structure - Introduction
Program
13/17
“Talking” to the Computer
The computer “understands”
machine language (‫)שפת מכונה‬.
Which is a mix of high and low
voltage signals passing through
transistors.
This language is also called binary code as
there are only two symbols in it.
How do we translate our C++ or Java code to
a language the computer understands?
Computer
Structure - Introduction
14/17
 After we have thought and came up with an idea.We need a
programmer. Ada Augusta King (1842), Countess of
Lovelace, translates Menabrea's pamphlet on the Analytical
Engine (the second version of Babbage’s Difference
Machine), adding her own notes, and becomes the world's
first programmer. The idea is expressed in some computer
language and entered into the computer using an editor.
 In order to “talk” to the computer we must send it electronic
signals. The easiest signals for the machine to understand
are on and off (in real life these signals are high voltage and
low voltage).
 Thus the computer’s alphabet is composed of two symbols 0
and 1. Any “words” composed with these 2 numbers are
called binary numbers.
 Computers are slaves to our commands and each individual
command is called an instruction.
 For instance the binary number 10001100100000 tells a
computer to add 2 numbers.
Assembly Language
The first programmers (back in the late 40s)
spoke to computers using binary numbers.
This, of course, was very tiring. So codes for
the instructions were created.
Soon a program was written that converted
these symbols to binary. This program was
called an assembler , it translated the
assembly language into binary.
 The next stage was to write in high-level
languages.
Computer
Structure - Introduction
15/17
 While writing in assembly code is much more productive
than writing in binary it is still unnatural. Each machine
instruction must be written on a single line, forcing the
programmer to think like a machine.
 So if it is possible to translate assembly language to binary,
what prevents us from writing in some higher level?
 The answer is: nothing. Although more complex than the
assembler this program called the compiler translated highlevel languages (like C++ or Java) to assembler.
 Today more than 99\% of all code is written using high-level
languages.
 So why bother with assembler, as we will during this course?
In order to understand the machine better we must know
how to “talk” to it in its native language.
An Example
High-level
language
program
(in C)
swap(int v[], int k)
{int temp;
temp = v[k];
v[k] = v[k+1];
v[k+1] = temp;
}
C compiler
Assembly
language
program
(for MIPS)
swap:
muli $2, $5,4
add $2, $4,$2
lw
$15, 0($2)
lw
$16, 4($2)
sw
$16, 0($2)
sw
$15, 4($2)
jr
$31
Assembler
Binary machine
language
program
(for MIPS)
Computer
00000000101000010000000000011000
00000000100011100001100000100001
10001100011000100000000000000000
10001100111100100000000000000100
10101100111100100000000000000000
10101100011000100000000000000100
00000011111000000000000000001000
Structure - Introduction
16/17
User code
compile
hello.cpp
link
hello.obj
hello.exe
link
I/O library
Computer
Structure - Introduction
io.lib
 The high level expression in C++:
A + B
 Will be translated to the assembly statement:
add A,B
 The assembler will then translate it into the binary
instruction:
1000110010100000
 The C++ source file, is compiled into assembly code.
 It is then linked with an external library, and transformed
into an executable file.
 The executable file can then be read by the computer.
OK, but how does it work?
So we have an executable file,
what does the computer do with it?
In a nutshell (‫ )על רגל אחת‬the file is loaded
from the hard disk into memory (by, how
obvious, the loader), the processor reads the
instructions one by one and executes them.
This idea of having programs kept as data in
memory is known as a Von Neumann
computer.
Computer
Structure - Introduction
17/17
 1943 - Work on ENIAC was started
in at the University of Pennsylvania,
with John Mauchly and J. Presper
Eckert responsible for its
implementation. The US Army liaison
(‫ ) קישור‬was Herman Goldstine.
 1944 - John von Neumann joined the ENIAC project. The
idea of storing programs as numbers was proposed.
 1945 - von Neumann wrote a memo proposing a storedprogram computer called EDVAC. Goldstine distributed the
memo, put von Neumann’s name on it and omitted Eckert’s
and Mauchly’s names.
 Most computer historians agree the von Neumann received
far more credit than he deserved.
 The most prestigious (‫ ) יוקרתי‬award in the field of Computer
Architecture is the Eckert-Mauchly award.