CHAPTER

01 Computer Architecture

1
2
3
 1.1 Introduction to Computer Architecture

A computer (or computer system) is a device

that receives and processes data according
to a set of instructions, and produces the
processed data as a result. Computer
architecture describes how a computer
is designed and built to function. It also
includes how the various parts of a computer
are designed, organised and connected.

ER MS
KEY T
Computer
A device that receives and processes data
according to a set of instructions and
produces the processed data.

Computer architecture
A description of how a computer is
designed and built to function, including
how its various parts are designed,
organised and connected.

Data (singular: datum)

Information that is used in a computer
program.

U
DID YO
?
KN OW
There are many ways to build a computer. While most modern computers
run on electricity, early computers did not use electricity at all. Instead,
they relied on mechanisms such as cranks, gears, pulleys and levers. These
mechanical computers used a completely different computer architecture
from the computers we are familiar with today.

For example, Figure 1.1 shows a toy called the Dig-Comp II that uses a set
of switches to represent data. These switches are connected to ramps. As
marbles roll down the ramp, they trigger the switches in different ways.
Because of this, the toy is able to perform computations such as addition,
functioning as a mechanical computer that works without electricity.
Figure 1.1 Digi-Comp II

4
 1.2 Units of Data

LEARNING OUTCOMES

1.1.1 Perform calculations using bits, bytes, kilobytes, kibibytes, megabytes, mebibytes,
gigabytes, gibibytes, terabytes, tebibytes, petabytes and pebibytes.

Since a computer’s function is to receive, process and

produce data, an important aspect of computer architecture ER MS
is to define how data is being represented. KEY T
Modern computers are digital devices evolved from machines Bit
that perform calculations on numerical data represented in A binary digit that takes on a
binary form. The smallest unit of data is a bit. A bit is a binary value of either 0 or 1.
digit that takes on a value of either 0 or 1. A computer stores
and processes all data using binary numbers that consist of Byte
these digits. A binary number made up of
eight bits.
A single bit would be too simple to represent complex data,
so we usually talk about data being represented as bytes
instead. A byte is a unit of data made up of eight bits.

Table 1.1 summarises some units of measurement of data in

order of increasing size.

Name of unit Symbol Size in bytes

kilobyte kB 1,000

kibibyte KiB 1,024

megabyte MB 1,0002 = 1,000,000

mebibyte MiB 1,0242 = 1,048,576

gigabyte GB 1,0003 = 1,000,000,000

gibibyte GiB 1,0243 = 1,073,741,824

terabyte TB 1,0004 = 1,000,000,000,000

tebibyte TiB 1,0244 = 1,099,511,627,776

petabyte PB 1,0005 = 1,000,000,000,000,000

pebibyte PiB 1,0245 = 1,125,899,906,842,624

Table 1.1 Units of measurement of data

5
 U
DID YO
?
KNO W
In the past, units of data such as “kilobyte”, “megabyte” and “gigabyte” were
based on powers of 1,024 (or 210) instead of the standard powers of 1,000 used in
the International System of Units (SI). However, since 1998, multiple standards
organisations have agreed that SI prefixes (i.e., “kilo”, “mega”, “giga”, etc.) should follow their standard
meanings while new binary prefixes (i.e., “kibi”, “mebi”, “gibi”, etc.) would be used to represent powers of
1,024.

Despite this change, you may still encounter the use of “kilobyte” and “megabyte” to represent 1,024 bytes
and 1,0242 bytes respectively, instead of the correct units “kibibyte” and “mebibyte”.

QUICK
E C K 1.2
CH
1. 1 kilobyte is the same size as 1,000 bytes. True or false?

2. Arrange the following units in increasing size: bit, byte, gibibyte, gigabyte, kibibyte, kilobyte

3. Convert the following amounts of data into bytes:

a) 2,026 kB b) 19 GiB c) 65 MB

4. Convert the following amounts of data into bits:

a) 2,026 kB b) 19 GiB c) 65 MB

1.3 Components of a Computer System

LEARNING OUTCOMES

1.1.2 Describe the function of key components of a computer system: its processor, main
memory and secondary storage.

1.1.3 Describe the function of data and address buses in reading from and writing to
memory.

1.1.4 Describe different input/output interfaces (USB, HDMI and PCI Express) in terms of
typical applications, connectors and speed.

1.1.5 Describe the use of magnetic, optical and solid-state media for secondary storage in
terms of durability, portability, typical capacities, cost and speed.

6
Most computer parts can be organised into the following
roles in Table 1.2. Figure 1.2 shows where the parts
performing these roles may be found inside the case of
a desktop computer.

U
DID YO
?
KNOW
f/g
Instead of having separate
parts, the key components of b
a computer in mobile devices
such as smartphones and
laptops are combined into a c
single part called a System d
on a Chip (SoC). This has the e
benefits of smaller size, lower
a
power consumption and,
often, improved performance.
However, SoCs are also
more complex and require
specialised expertise to
design and manufacture.

Figure 1.2 Components in a desktop computer case

Role Description

(a) Processor Processes data and follows instructions

(b) Memory Stores data and instructions temporarily for immediate use by
the processor

(c) Secondary storage Stores large amounts of data that will not be lost when the
power supply is interrupted

(d) Data bus Transports data between memory and processor; bi-directional

(e) Address bus Transports required memory location from processor to

memory; uni-directional

(f) Input Data or instructions that the computer receives

(g) Output Intermediate or final results produced by the computer; usually

in the form of processed data

Table 1.2 Inside the computer

7
 1.3.1 Processor

You have learnt that computers need to

process data and follow instructions. The
computer part that does this is called the
processor or central processing unit (CPU).
It is usually a complex circuit made of many
components compressed into a square or
rectangular package. You will learn more
about the components that make up this
circuit in Chapter 3. Figure 1.3 A processor package

When a processor ‘runs’ or ‘executes’ instructions,

MS
EY TER
it follows or performs instructions. A series of
K instructions is called a program, or simply software.

Central processing unit (or processor) Often, a processor’s speed is described by the number
The part of the computer that processes of instructions that the processor can perform in one
data and follows instructions. second. For instance, a 1 MHz (megahertz) processor
can perform one million instructions per second while
Execute a 1 GHz (gigahertz) processor can perform one billion
To follow or perform an instruction. instructions per second. In general, the larger the
number and unit, the faster and more powerful the
Software processor.
A set of instructions to perform specific
tasks on a computer. There are also “multi-core” processors that contain
multiple processing units inside a single package.
For instance, a “dual-core” package has two such
processing units while a “quad-core” package has
four. These “multi-core” processors can perform more
than one instruction at the same time, and thus are
more powerful than “single-core” processors.

U
DID YO
?
KN OW
Many computers also include a specialised computer
part called a Graphics Processing Unit (GPU) that is
more efficient at performing the calculations needed
to produce images than a CPU. By relieving the CPU
of tasks related to 2D and 3D graphics, GPUs help to
improve the overall performance of a system.

Since graphics often involve performing many

calculations simultaneously, GPUs have also become
useful in fields such as Artificial Intelligence that
require performing repetitive calculations on a large
amount of data. You will learn more about this topic
in Chapter 14.

8
 1.3.2 Main Memory

Whenever a computer needs to store data, it

uses memory.

There are different types of memory.

Random access memory (RAM) or main
memory is a type of memory where data
and instructions are stored temporarily so Figure 1.4 Example of a RAM card
that they can be quickly accessed by the
processor when needed. For instance, when
an application is started, its instructions
may be loaded into the RAM. Data stored
U
on the RAM can be easily changed and is DID YO
?
KNOW
also volatile, which means that it is lost Read-only memory (ROM)
once the power supply to the computer is is another type of memory
interrupted. that stores data and
instructions that rarely
The word “memory” usually refers to RAM need to change or would be needed for a computer to
instead of other types of memory. Most start up. Data stored on ROM cannot be easily changed
desktop computers have RAM located on and it is retained even when the power supply is
removable cards so that the amount of switched off. This makes it ideal for storing critical
RAM in the computer can be upgraded instructions that are required to start the computer
easily. Other kinds of computers, however, before other data can be loaded into the RAM.
may have only a fixed amount of RAM that
cannot be changed.

Physically, memory exists in different forms, but in general, it consists of many switches arranged in a fixed
order. Each switch can store one bit of data based on whether it is ON or OFF (i.e., ON = 1, OFF = 0). Since a
byte consists of eight bits, we usually look at eight consecutive switches at a time, as shown in Figure 1.5.

In Figure 1.5, big physical switches are used to represent memory, but in most computers, memory is made
up of very small electronic switches that take up much less space.

Usually, the position of each byte is represented by a number called an address. This is a number that
allows a computer to quickly find those switches again if it needs to read or change the data that is stored.
This is similar to how a unit number is used to identify the location of each residence in an HDB block.

ERMS
KEY T
Address
A number that is used to locate a byte in
memory.

Memory (Main Memory)

A device that is used to store data and
instructions temporarily for immediate
use by the processor.

Volatile
Lost when the power supply is interrupted.
Figure 1.5 Memory as a collection of switches in a fixed order

9
 1.3.3 Secondary Storage

ERMS
KEY T
Secondary storage is a way of storing large
amounts of data that will not be lost when the
power supply is interrupted. Compared to RAM, Secondary storage
secondary storage is usually cheaper and able to A way of storing large amounts of data
store much more data. It is also non-volatile, so the that will not be lost when the power
data that is stored remains there even without a supply is interrupted.
power supply. This makes secondary storage ideal
for physically transporting data from one computer Hard disk (or hard drive)
to another. On the other hand, secondary storage Secondary storage where data is stored
is usually much slower in speed compared to RAM. on rigid rotating disks coated with a
magnetic material.
The word “storage” usually refers to secondary
storage.

The processor usually does not access data in

secondary storage directly. Instead, any data in
secondary storage that the processor needs might
be copied to the RAM first.

1.3.3.1 Choosing Storage Media for Secondary Storage

There are many types of storage media available. When deciding on the best type of storage media to use,
these factors should be considered:

Figure 1.6 Factors to consider when choosing storage media

10
 1.3.3.2 Types of Storage Media

Table 1.3 summarises three main types of storage media in terms of durability, portability, typical capacities,
cost and speed.

Type Magnetic Optical Solid-State

Description Data is stored on a Data is stored as Data is stored in

magnetic material that very small pits or electronic circuits that
can be read or written by indentations that can have no moving parts
a magnetic “head” be read or written by a
laser Example: memory
Example: hard disk cards
Example: digital
versatile disc (DVD)

Example: magnetic tape

Example: solid-state
drive

Durability More vulnerable to Vulnerable to damage Most durable;

damage from from Most resistant to
• magnetic fields, • scratches, and • impact, and
• heat, • natural deterioration • temperature changes
• impact, and over time
• natural deterioration More resistant to
over time • heat
• impact

Portability Heavier and bulkier Portable due to small Portable due to small
than optical and solid- size and light weight size and light weight
state media

Typical Up to TBs of data Up to GBs of data Up to TBs of data

Capacities

Cost per Lowest Lower cost than solid- Highest

GB state media but higher
cost than magnetic
media

Speed Slower than solid-state Slower than solid-state Faster than magnetic
media media and optical media

Table 1.3 Three main types of storage media

11
 1.3.4 Data and Address Buses

Besides storing data, computers also use buses

to transport data from one part of the computer
to another. A bus is a collection of wires that
serves as a “highway” for data to travel on. It can
be made of either physical wires or conductive
lines printed on a circuit board. Figure 1.7 Example of a computer bus made of physical wires

Front of circuit board Back of circuit board

Conductive lines
(including data and address buses)

Figure 1.8 Example of computer buses made of conductive lines

ER MS
KEY T Two important buses that transport data between
the processor and memory parts of a computer are
Bi-directional the data bus and the address bus:
Able to work in two directions, to and fro.
1. The data bus transports data that is going to be
Data bus processed to the CPU, and transports data that has
A bus that is used to transport data between already been processed from the CPU. The data bus
memory and the processor. is bi-directional because data can be sent in both
directions between the processor and memory.
Address bus
A bus that is used to specify memory address 2. The address bus specifies memory address
information. information. When the processor reads from or
writes to the memory (RAM), the relevant address
Uni-directional information is provided on the address bus. The
Able to work in one direction only. address bus is uni-directional because address
information is always sent in one direction only,
that is, from the processor to the memory.

For instance, to read data from the memory, the processor requests the data, and the address bus transports
the requested data’s address to the RAM. A copy of the requested data is then sent from the RAM back to
the processor via the data bus. This is illustrated in Figure 1.9, which uses bits for data and simple numbers
for addresses.

12
 0 01001011
Data Bus 1 10100101
01100101 2 01100101
3 10110010
4 01100100
2
5 01110101
6 11001100
and so on
...

Figure 1.9 How the data bus and address bus are used to read from memory

When writing data to the memory, the processor uses both the data bus and address bus at the same
time to transport the data for writing as well as the destination address to the RAM. The RAM then sets the
switches at the destination address according to the data received via the data bus, as illustrated in Figure
1.10.

Before writing to memory After writing to memory

0 01001011 0 01001011
1 10100101 1 10100101
11111111 2 01100101 2 01100101
Data Bus 3 10110010 3 10110010
4 01100100 4 11111111
4
5 01110101 5 01110101
Address Bus
6 11001100 6 11001100
and so on and so on
... ...

Figure 1.10 How the data bus and address bus are used to write to memory

As shown by the arrows in Figure 1.9 and Figure 1.10, while the address bus is always used to transport
information from the processor to the memory, the data bus can transport information in either direction.

1.3.5 Input and Output Interfaces

In computer architecture, input refers to data or instructions that the computer receives for processing
while output refers to any intermediate or final results produced by the computer in the form of processed
data. Examples of input are words entered using a keyboard, pictures taken by a digital camera, and
movement instructions entered using a mouse. Examples of output are images displayed on a screen,
sounds played on a speaker and even sculptures printed using a 3D printer.

Often, input is received from, and output is sent to hardware devices:

• An input device is a hardware device that allows users to enter data and instructions into a computer.
Examples of input devices are keyboards, mice, scanners, touch screens and microphones.
• An output device is a hardware device used to display, project or print processed data from a
computer so it can be used or understood by people using the computer. Examples of output devices
are monitors, speakers and printers.

13
 A computer usually has multiple input and output devices connected to it, as seen in Figure 1.11.

ER MS
KEY T
Input (computer architecture)
Data or instructions that the computer
receives for processing

Output (computer architecture)

Intermediate or final results produced
by the computer in the form of processed
data

Hardware
Physical components of a computer

Input device
A hardware device that allows users
to enter data and instructions into a
computer

Output device
A hardware device used to display, project
or print processed data from a computer

Figure 1.11 Examples of connected input and output devices

The method by which a computer connects to an input or output device is called an input/output interface.
Different input/output interfaces have varied uses, physical connectors and maximum data transmission
speeds as new versions of each interface are introduced over time. Table 1.4 summarises this information
for three common input/output interfaces:

Interface Typical Applications Connectors Maximum Speed

Universal For powering and/ USB Type A USB Micro USB 2.0: 480 Mbit/s
Serial Bus or communicating USB 3.2: 20 Gbit/s
(USB) with external devices USB4: 80 Gbit/s
(e.g., printers, mice,
keyboards)
USB Type B USB Type C

USB Mini

14
 Interface Typical Applications Connectors Maximum Speed

High- For delivering HDMI Standard HDMI 1.3-1.4b:

Definition audio/video data to 10.2 Gbit/s
Multimedia compatible devices HDMI 2.0-2.0b:
Interface (e.g., monitors, TVs) 18 Gbit/s
(HDMI) HDMI 2.1:
48 Gbit/s

HDMI Mini

HDMI Micro

Peripheral For communicating PCI-e x1 Increases with

Component with internal number of lanes
Interconnect expansion cards (e.g., (up to x16)
Express (PCI graphics cards) PCI-e x4
Express) Max speed per lane
PCI Express 5.0:
PCI-e x8 4 GB/s
PCI Express 6.0:
8 GB/s
PCI-e x16 PCI Express 7.0:
16 GB/s

Table 1.4 Three common input/output interfaces

Unlike USB and HDMI which typically have connectors on

the side or back of a computer, PCI Express connectors are
typically located inside a computer on its motherboard.
Each connector has a number of lanes for transferring
ERMS
KEY T
data and the number of lanes determines the length
of the connector as well as the maximum speed for
that connector. By convention, the number of lanes is
described using the abbreviation “x” followed by the Lane (PCI Express)
number of lanes (e.g., x1, x4, x8 and x16). An interface for transferring data
between a computer and an expansion
While it is not necessary to memorise the maximum card.
speed of each interface, it is useful to recognise that
the maximum speed of each interface roughly doubles Motherboard
or more with each major revision and that the internal The main circuit board in a computer
PCI Express interface is generally faster than the external that connects all the components
USB and HDMI interfaces. By convention, the speed of together.
PCI Express lanes is expressed in GB/s instead of gigabits
per second (Gbit/s) and that 8 Gbit/s = 1 GB/s.

15
 QUICK
C K 1. 23
CHE
1. Decide if each of the following statements on memory and storage are true or false.
a) Both memory and storage may contain data.
b) The contents of memory are retained when a computer restarts.
c) The contents of storage are retained when a computer restarts.
d) The typical cost per GB is cheaper for memory than for storage.
e) The typical capacity is higher for memory than for storage.

2. For each of the following situations, suggest an appropriate secondary storage medium that should be
used and give a reason for your choice:
a) Storing terabytes of video files for backup
b) Keeping an emergency copy of some important files in your wallet

3. The following table shows the first 4 bytes of a computer’s memory:

Address Contents
0 00000000
1 11110000
2 00110011
3 00001111

The processor then executes an instruction to write the 8 bits 11111111 to address 3.
a) To execute the instruction, describe what information must be sent on the address bus and in what
direction.
b) Show the resulting first 4 bytes of memory after the instruction is executed.

4. Memory addresses can also be read from or written to memory as data. True or false?

5. Which of the three I/O interfaces (USB, HDMI and PCI Express) are suitable for directly connecting:
a) a wireless network adapter? b) an internal sound card?
c) a pair of audio speakers? d) a microphone?

W
REVIE
E STI ON
Q U
1. Describe the function of each of the following components:
a) Processor b) Main memory c) Secondary storage

2. Calculate the number of bits (not bytes) in the following amounts of data:
a) 8 TB b) 0.125 MB c) 0.125 MiB

3. The position of each byte in memory is called its address.

a) Calculate the number of addresses needed for 64 KiB of memory.
b) The memory in a computer has 256 possible addresses. Calculate the total size of the computer’s memory
in bits (not bytes).

4. A hard disk is used as secondary storage for a laptop.

a) A data bus connects the laptop’s processor directly to its hard disk. True or false?
b) The laptop’s owner replaces the hard disk with a solid-state drive. Identify two improvements that the
owner is likely to experience.

16
ANSWER

Pg. 6-Quick Check 1.2

1. True. However, “kilobyte” is still often confused with “kibibyte”, which is 1,024 bytes.

2. bit, byte, kilobyte, kibibyte, gigabyte, gibibyte

3. Show workings for each conversion.

a) 2,026 kB = 2,026 × (1,000 bytes) = 2,026,000 bytes
b) 19 GiB = 19 × (1,0243 bytes) = 20,401,094,656 bytes
c) 65 MB = 65 × (1,0002 bytes) = 65,000,000 bytes

4. Note that the answers below are just the answers for Q3 multiplied by 8 because 1 byte is the same as 8 bits.
a) 2,026 kB = 2,026 × (1,000 × 8 bits) = 16,208,000 bits
b) 19 GiB = 19 × (1,0243 × 8 bits) = 163,208,757,248 bits
c) 65 MB = 65 × (1,0002 × 8 bits) = 520,000,000 bits

Pg. 16-Quick Check 1.3

1. State true/false for each part (explanations are not required).
a) True. b) False. c) True.
d) False. (Storage usually is cheaper.)
e) False. (Storage usually has higher capacity.)

2. For each part, give a suggestion and state a reason.

a) A hard disk. It is the most common and affordable device with storage capacity available in terabytes.
b) A memory card. It is the only device small and flat enough to fit in a wallet. A memory card would also be more
durable and resistant to impacts that could occur in a wallet.

3. Answers as follows:
a) The address 3 must be sent on the address bus. It is sent from the processor to memory.
b) Only the contents at address 3 is changed:
Address Contents
0 00000000
1 11110000
2 00110011
3 11111111

4. True. Memory addresses are just numbers so they can also be treated as data.

5. Suitable I/O interfaces:

a) USB and PCI Express b) PCI Express only
c) USB and HDMI d) USB only

17
 ANSWER

Pg. 16-Review Question

1. Based on definitions of each key term:
a) Processor: processes data and follows instructions.
b) Main memory: stores data and instructions temporarily for immediate use by the processor.
c) Secondary storage: a way of storing large amounts of data that will not be lost when power supply is
interrupted.

2. Show workings for each conversion.

a) 8 × 1,000 × 1,000 × 1,000 × 1,000 × 8 = 64,000,000,000,000 bits
b) 0.125 × 1,000 × 1,000 × 8 = 1,000,000 bits
c) 0.125 × 1,024 × 1,024 × 8 = 1,048,576 bits

3. Show relevant workings for each part.

a) 64 × 1,024 = 65,536 bytes, which need 65,536 addresses
b) Each address represents one byte, so if there are only 256 possible addresses, there are only 256 bytes of
memory. Hence, total size of memory = 256 × 8 = 2,048 bits

4. Answer for each part as follows:

a) False. The data bus connects the processor to memory only and not to secondary storage such as hard disks.
b) Two improvements (accept any possible answers):
• Less vulnerable to damage from physical impact or external magnetic fields
• Faster access speeds
• Lighter weight
• Less noise and power use due to no moving parts (longer laptop battery life)

18