Module 5Central Processing Unit
1. Binary representation of data
2. The components of the CPU
3. CPU and Instruction set
4. Important features of CPUs
2
Data representation• All modern digital equipment work on binary digits,
0, 1, based on the two measurable/recogniable states: such as circuitry voltage, high voltage (ON): 1 Low voltage (Off) : 0
• Why? Simple, reliable, and available technology
• The fundamental component made of silicon: Transister
• Logic gates AND, OR, NOT can be constructed from transistor
3
Data representation• All modern digital equipment work on binary digits,
i.e., 0, 1, based on the two states of a circuitry, e.g.High voltage (ON): 1 Low voltage (Off) : 0
• Why? Simple, reliable, and available technology
• CPU can only understand 0 and 1
• Data must be represented in binary form before it is processed by a CPU – How to representation number?
• Binary (base 2) number, containing only two digits, 0 and 1– How to represent characters?
• English text for example
4
Binary number
• In decimal (base 10) we indicate place values by the order of digits– 135 = (1*102) + (3*101) + (5*100)
• In binary only 0’s and 1’s exist we do the same thing but since we use only two digits the place values change.– 10011 = (1*24)+(0*23)+(0*22)+(1*21)+(1*20)
= 16 + 0 + 0 + 2 + 1 = 19
5
Conversion between binary and decimal 10112 = (1x20)+(1x21)+(0x22)+(1x23)
= 1+2+0+8
= 11
Decimal to binary: repeatedly divide the quote by 2, read the remainder backward
115 12 11 00 1
1110=10112
Readup
199 14 12 01 0
1910 = 100112
Readup
0 1
6
One binary digit can represents two numbers, 0, 1
Two binary digits can represent four numbers 00, 01, 10, 11
0 1 2 3
Three binary digits can represent ___numbers
Fixed length binary representation
7
Example of 4 binary digits numbers
8
Binary number facts
Bits Range
8 0 to 28-1=255
16 0 to 216-1= 65 535
32 0 to 232-1= 4 294 967 295
64 0 to 264-1=18 446 744 073 709 551 615
It takes more digits to represent a number in binary than it does in decimal
9
Binary number operations
Add, subtract, multiply, division
Example of add
1101 + 1001
1101
1001
+_____
10110
10
Binary representation for characters
• Coding schemes provide a common way of representing a character of data
• Common Schemes– ASCII, 7 bits– Stands for American Standard Code for Information
Interchange– Most widely used standard– Used on virtually all personal computers
11
ASCII coding scheme• 7 bits, 127 different symbols• Examples
Dec Binary Symbols
048 0110000 0 049 0110001 1 050 0110010 2 051 0110011 3 052 0110100 4 053 0110101 5 054 0110110 6 055 0110111 7 056 0111000 8 057 0111001 9 058 0111010 : (colon) 059 0111011 ; (semi-colon) 060 0111100 < (less than) 061 0111101 = (equal sign) 062 0111110 > (greater than) 063 0111111 ? (question mark) 064 1000000 @ (AT symbol)
065 1000001 A065 1000001 A 066 1000010 B066 1000010 B 067 1000011 C067 1000011 C 068 1000100 D068 1000100 D 069 1000101 E069 1000101 E 070 1000110 F070 1000110 F 071 1000111 G071 1000111 G 072 1001000 H072 1001000 H 073 1001001 I073 1001001 I 074 1001010 J074 1001010 J 075 1001011 K075 1001011 K 076 1001100 L076 1001100 L 077 1001101 M077 1001101 M 078 1001110 N078 1001110 N 079 1001111 O079 1001111 O 080 1010000 P080 1010000 P 081 1010001 Q081 1010001 Q 082 1010010 R082 1010010 R 083 1010011 S083 1010011 S 084 1010100 T084 1010100 T 085 1010101 U085 1010101 U 086 1010110 V086 1010110 V 087 1010111 W087 1010111 W 088 1011000 X088 1011000 X 089 1011001 Y089 1011001 Y 090 1011010 Z090 1011010 Z
12
Unicode
• Designed to accommodate alphabets of more than 256 characters
• Uses 16 bits to represent one character– 65,536 possible values
• Requires twice as much space to store data
14
Word• Word is the number of bits the CPU processes
as a unit.
• A word usually consists of several bytes. It is CPU dependent. The larger the word, the more powerful the computer.
• We say a 32 bits processor. Here 32 is the size of the word. It is equal to the size of a register. in length
15
The units for information storage sizes
• One kilobyte (1 KB) = 1024 (210) bytes = 213 bits– Memory capacity of older personal computers
• One Megabyte (1 MB) = 220 bytes ≈ 106 bytes (one million bytes)– Personal computer memory – Portable storage devices (diskette, CD-ROM)
• One gigabyte (1 GB) = 230 bytes ≈ 109 bytes (one billion bytes) – Storage devices (hard drives)– Mainframe and network server memory
• One terabyte (1 TB) = 240 bytes ≈ 1012 bytes (one trillion bytes)
– Storage devices on very large systems
16
Central Processing Unit (CPU)
• Consists of complex set of electronic circuitry • Executes stored program instructions• Three components
– Registers– Control unit– Arithmetic/Logic Unit
(ALU) Control Unit
ALU
Registers
Memory
CPU
17
Registers
• Registers: temporary place for instructions and data. All instructions and data must be placed in registers before being executed and processed
• High-speed temporary storage areas– Storage locations located within the CPU
• Work under direction of control unit– Accept, hold, and transfer instructions or data– Keep track of where the next instruction to be
executed or needed data is stored
18
Control unit
• Control unit: directs the computer system to execute stored program instructions
• Must communicate with memory and ALU
• Storage operations– Moving data/instructions from/to register to/from
registers – Sends data and instructions from secondary storage
to memory as needed
19
ALU
• Executes all arithmetic and logical operations– Arithmetic operations
• Addition, subtraction, multiplication, division
– Logical operations• Compare numbers, letters, or special characters• Tests for one of three conditions
– Equal-to condition, Less-than condition, Greater-than condition
– Branch to different instructions
20
Memory
• Also known as primary storage and main memory– Often expressed as random-access memory
(RAM)– Not part of the CPU
• Holds data and instructions for processing
• Stores information only as long as the program is in operation
21
The CPU and memory
• CPU cannot process data from disk or input device– It must first reside in memory– Control unit retrieves data from disk and moves it into
memory
• Items sent to ALU for processing– Control unit sends items to ALU, then sends back to
memory after processing
• Data and instructions held in memory until sent to an output or storage device or program is shut down
22
Data storage and the CPU
• Two types of storage:– Primary storage (memory)
• Stores data temporarily• CPU refers to it for both program instructions and
data
– Secondary storage• Long-term storage• Stored on external medium, such as a disk
23
How the CPU executes instructions
• Four steps performed for each instruction– Machine cycle: the amount of time needed
to execute an instruction– Personal computers execute in less than
one millionth of a second– Supercomputers execute in less than one
trillionth of a second
• Each CPU has its own instruction set– those instructions that CPU can understand
and execute
24
The machine cycle
• The time required to retrieve, execute, and store an operation
• Components– Instruction time– Execution time
• System clock synchronizes operations
25
Instruction time
• Also called I-time
• Control unit gets instruction from memory and puts it into a register
• Control unit decodes instruction and determines the memory location of needed data
26
Execution time
• Control unit moves data from memory to registers in ALU– ALU executes instruction on the data
• Control unit stores result of operation in memory or in a register
27
Microprocessor
• Central processing unit etched on silicon chip
• Contain hundreds of millions of tiny transistors
• Key components:– ALU– Registers– Control unit– System clock
28
Transistors• Electronic switches that may or may not allow
electric current to pass through– If current passes through, switch is on, representing a
1 bit– Otherwise, switch is off, representing a 0 bit
• The size of a transistor– New technology makes the size of a transistor smaller
and smaller! – Current: 0.13 micro meter or 130 nano meter
Moving to 0.09 micro meter 90 nano meterNext a few years: 60 nano meters
29
Types of chips
• Intel makes a family of processors– Pentium III and Pentium4 processors in most PCs– Celeron processor sold for low-cost PCs– Xeon and Itanium for high-end workstations and
network servers
• Other processors– Cyrix and AMD make Intel-compatible
microprocessors– PowerPC chips used primarily in Macintosh
computers– Compaq’s Alpha microprocessor used in high-end
servers
30
Computer processing speeds
• Instruction speeds measured in fractions of seconds– Millisecond: one thousandth of a second– Microsecond: one millionth of a second– Nanosecond: one billionth of a second
• Modern computers have reached this speed
– Picoseconds: one trillionth of a second
31
Microprocessor speeds
• Measure of system clock speed – How many electronic pulses the clock produces per
second (clock frequency)– Usually expressed in frequency
• Gigahertz (GHz), Megahertz (MHz), Kilohertz (KHz)
– Or how much time for each cycle: clock cycle time clock cycle time = 1 / clock frequency
• Second, millisecond, microsecond, nanosecond, picoseconds
• Comparison of clock speed only meaningful between identical microprocessors
32
Other performance measures
• Millions of Instructions per Second (MIPS)– High-speed personal computers can perform
over 500 MIPS– Typically a more accurate measure of
performance than clock speed
• Megaflop: one million floating-point operations– Measures ability of computer to perform
complex mathematical operations
33
Cache
• A temporary storage area– Speeds up data
transfer within computer
• Processor cache
Control Unit
ALU
Registers
Memory
CPU
Cache L1
Cache L2
34
Processor cache• A small block of high-speed memory
– Stores most frequently and most recently used data and instructions
• Microprocessor looks for what it needs in cache first– Transferred from cache much faster than from memory– If not in cache, control unit retrieves from memory
• The more cache “hits” the faster the system performance
• Internal (Level 1) cache built into microprocessor– Fastest access, but highest cost
• External (Level 2) cache on separate chip– Incorporated into processor on some current microprocessors
Top Related