IGCSE Computer Science: Hardware — Revision Notes

Condensed for the final weeks. For the full explanation, use the Hardware study guide.

3.1 Computer architecture — the FDE cycle

Component	Role
ALU	Performs arithmetic/logic operations
CU	Controls the cycle’s timing and sequencing
PC	Holds the address of the next instruction
MAR	Holds the address currently being accessed
MDR	Holds data/instructions moving to/from memory
CIR	Holds the instruction currently being decoded
ACC	Holds the result of ALU operations

Buses carry information between components: address bus (where), data bus (what), control bus (signals). Know the cycle in order — fetch (instruction from RAM to CIR via MAR/MDR), decode (CU interprets it), execute (ALU/CU carry it out).

Core/cache/clock: more cores, larger cache, faster clock all improve performance — but be ready to say why, not just list them. Embedded system = one dedicated function (a washing machine controller); a PC is general-purpose.

3.2 Input, output and sensors

Category	Named examples
Input	Barcode scanner, digital camera, keyboard, microphone, optical mouse, QR scanner, touch screen (resistive/capacitive/infra-red), 2D/3D scanner
Output	Actuator, DLP/LCD projector, inkjet/laser printer, LED/LCD screen, speaker, 3D printer
Sensor	Acoustic, accelerometer, flow, gas, humidity, infra-red, level, light, magnetic field, moisture, pH, pressure, proximity, temperature

Scenario questions (“a greenhouse needs to monitor…”) test whether you can match a sensor to what it measures — build a quick-recall table of sensor → measured quantity rather than memorising the list alphabetically.

3.3 Data storage

Type	Accessed by CPU directly?	Volatile?	Examples
Primary (RAM)	Yes	Yes	Current programs/data
Primary (ROM)	Yes	No	Startup instructions
Secondary — magnetic	No	No	HDD (platters, tracks, sectors, electromagnets)
Secondary — optical	No	No	CD/DVD/Blu-ray (laser, pits/lands)
Secondary — solid-state	No	No	SSD/SD/USB (NAND/NOR, transistors)

Virtual memory: pages swapped between RAM and disk when RAM is full. Cloud storage: remote access vs needing physical local servers — know one advantage and one disadvantage of each side.

3.4 Network hardware

• NIC — required to access a network.
• MAC address — fixed at manufacture, hexadecimal, manufacturer code + serial code.
• IP address — allocated by the network, static or dynamic; know IPv4 vs IPv6 differences (address length/format).
• Router — directs data to its destination, can assign IP addresses, connects a local network to the internet.

Links to Topic 2 (Data transmission): a router directing packets builds directly on packet-switching — revise both together.

Worked example: tracing an instruction through the FDE cycle

A common exam format describes a single instruction and asks you to trace its path through the cycle. Work through it in this order every time:

1. FETCH:   address of next instruction copied from PC into MAR
            instruction itself copied from RAM into MDR, then into CIR
            PC incremented to point at the following instruction
2. DECODE:  CU interprets the instruction held in CIR
3. EXECUTE: ALU performs the operation; result stored in ACC
            (or written back to RAM if the instruction requires it)

Practising this trace with different named registers filled in for each stage is worth more than memorising the register list in isolation, since exam questions usually ask you to identify which register or bus is active at a specific stage, not to recite the whole cycle from memory.

Worked example: matching sensors to a real scenario

A car’s automatic parking system needs to detect nearby obstacles and adjust its dashboard lighting to match ambient conditions. Identify suitable sensors.

Detects nearby obstacles:       Proximity sensor (or acoustic sensor,
                                 depending on the technology used)
Adjusts dashboard lighting:     Light sensor

As with all sensor-scenario questions, the mark is earned by naming the sensor and stating what physical quantity it detects — not simply recognising that “a sensor” is involved.

Why RAM and ROM are both needed

A frequent exam question asks why a computer needs both RAM and ROM rather than just one. RAM is fast and directly accessible, but volatile — its contents vanish when power is removed, so it cannot hold the instructions needed to start the computer before an operating system even exists in memory. ROM is also directly accessible but non-volatile, so it permanently stores the bootstrap instructions that begin the startup process. Once started, the operating system and running programs are loaded into RAM for speed, while ROM’s contents remain unused again until the next power-on — the two types of primary storage serve genuinely different purposes rather than being interchangeable options.

Exam traps

• Naming FDE registers without being able to trace a single instruction through fetch → decode → execute in order.
• Confusing a sensor’s name with what it actually measures under scenario pressure.
• Saying “RAM is used for storage” without distinguishing primary (direct CPU access) from secondary (not directly accessed).
• Mixing up MAC address (fixed, hardware) with IP address (allocated, network-level, can change).
• Listing cloud storage’s benefits without a matched limitation (accessibility vs. reliance on connectivity/host infrastructure).

Self-test

1. Name the four buses/units involved in the FDE cycle’s “fetch” stage and what each does.
2. What distinguishes primary from secondary storage?
3. Give one advantage and one disadvantage of cloud storage vs local storage.
4. What is the difference between a MAC address and an IP address?
5. Why is embedded system correctly distinguished from a general-purpose computer?

Answers: 1. PC (holds next instruction’s address), MAR (holds the address being accessed), MDR (holds the data/instruction being transferred), address/data/control buses (carry the address, data and signals between components). 2. Primary storage is directly accessed by the CPU; secondary storage is not and is used for more permanent data. 3. Advantage: remote accessibility from any location; disadvantage: dependence on physical servers/infrastructure to host it and a network connection to access it. 4. A MAC address is fixed at manufacture and identifies the specific network hardware; an IP address is allocated by the network, can be static or dynamic, and identifies a device’s location on that network. 5. An embedded system performs one dedicated function (e.g. a vending machine controller), while a general-purpose computer such as a PC runs many different programs and functions.