Chapter 03

Hardware

Cambridge O Level Computer Science 2210 Topic 3 revision chapter covering computer architecture (the role of the CPU, what a microprocessor is, the purpose of every component of a Von Neumann CPU including the control unit, arithmetic logic unit, program counter, memory address register, memory data register, current instruction register and accumulator, the address, data and control buses, the full fetch-decode-execute cycle, how the number of cores, the size of the cache and the speed of the clock affect CPU performance, the purpose and use of an instruction set, and the purpose, characteristics and typical devices of embedded systems), input and output devices (what input and output devices are and why they are required, the required list of input devices from barcode scanners to 3D scanners, the required list of output devices from actuators to 3D printers, and the fourteen required sensor types with the physical property each one measures and how to choose the most suitable sensor for a context), data storage (primary storage and the different roles of RAM and ROM, secondary storage, the operation of magnetic, optical and solid-state flash storage with examples of each, how virtual memory is created and used and why it is necessary, and what cloud storage is together with the advantages and disadvantages of cloud storage compared with local storage) and network hardware (why a computer needs a network interface card, the purpose and structure of a MAC address, the purpose of an IP address and the difference between static and dynamic allocation, the characteristics of and differences between IPv4 and IPv6, and the role of a router in a network).Show more

Syllabus 22102026-2028

Core Revision Module

Recommended Next

Revision & Practice Book

Interactive revision notes with exam tips and worked examples for this chapter.

Full syllabus coverage•Exam tips

Open Notes

Practice & Resources

2 tools

Paper 1 · Multiple Choice

Multiple-choice questions with explanations.

Start

Paper 2 · Theory

Mark-scheme-style structured questions.

Start

Chapter overview

A summary of this Computer Science chapter — open a section to read it. The full notes, worked examples and practice questions are in the study modules above.

What is Hardware about?

Topic 3 is one machine seen from four distances. Up close, the CPU repeats a single loop — fetch an instruction, decode it, execute it — using named registers and three buses. Step back and the CPU is only useful because input devices and sensors turn physical things into data and output devices and actuators turn data back into physical things. Step back again and nothing survives a power cut unless storage holds it: RAM while the machine runs, secondary or cloud storage afterwards. Step back once more and the machine needs network hardware — a NIC with a MAC address, an IP address and a router — before it can be found and reached by anything else.

Chapter 3 is Topic 3, which belongs to Paper 1: Computer Systems. Paper 1 is a written paper lasting 1 hour 45 minutes and carrying 75 marks. It contains short-answer and structured questions set on Topics 1–6, all questions are compulsory, and calculators are not permitted. Those five facts come from the assessment overview of the 2026–2028 syllabus.

A register is a very small, very fast store inside the CPU that holds one value at a time. The five you must know split cleanly into two families. PC and MAR hold a “where” — they contain memory addresses. MDR, CIR and ACC hold a “what” — they contain actual instructions or data values. Almost every register mark lost in Paper 1 is lost by putting a register in the wrong family.

A bus is a set of parallel connections that carries signals between the CPU, memory and input/output components. There are three, and the quickest way to keep them apart is one word each: the address bus carries WHERE, the data bus carries WHAT, and the control bus carries WHEN and WHICH WAY.

Three separate things affect how much work a CPU gets through. The clock sets how many cycles happen per second, so a higher clock speed can allow more processing operations per second. The number of cores sets how many instruction streams the CPU can work on at the same time. The size of the cache sets how much frequently used data and instructions can be kept close to the CPU, so it has to wait for the slower RAM less often. Each one can improve performance; none of them guarantees it on its own.

An instruction set is the complete list of all the commands that a particular CPU can process, and those commands are machine code. It is a property of the processor design, not of any one program: every program that runs on that CPU must, in the end, be expressed using instructions from its set.

Key ideas to remember

The one sentence that carries this chapter: data enters, is stored, is fetched and executed one instruction at a time by named components, and leaves again as output — so every answer should name a component, state what it holds or carries, and only then describe what happens.
One habit to carry into the exam: before writing any answer in Topic 3, name the component. “The control unit sends…”, “the MAR holds…”, “a level sensor measures…”. Almost every mark in this topic is attached to a name.

What you need to be able to do

3.1.1a Explain the role of the central processing unit in a computer.
3.1.1b Explain what is meant by a microprocessor.
3.1.2a State the purpose of each component in a CPU that has a Von Neumann architecture — the ALU and control unit, the PC, MAR, MDR, CIR and ACC registers, and the address, data and control buses.
3.1.2b Describe the fetch–decode–execute cycle, including the role each component plays in it.
3.1.3 Explain what a core, a cache and a clock are, and how each can affect the performance of a CPU.
3.1.4 Explain the purpose and use of a CPU instruction set.
3.1.5 Describe the purpose and characteristics of an embedded system and identify devices that commonly contain one.
3.2.1 Explain what an input device is and why one is required, across the required list of devices.
3.2.2 Explain what an output device is and why one is required, across the required list of devices.
3.2.3a Explain what a sensor is and what sensors are for.
3.2.3b Identify the type of data each of the fourteen listed sensors captures, state when each would be used, and select the most suitable sensor for a given context.
3.3.1 Explain what primary storage is, including the roles of RAM and ROM, why a computer needs both, and the differences between them.
3.3.2 Explain what secondary storage is and why it is necessary.
3.3.3 Describe the operation of magnetic, optical and solid-state (flash memory) storage and give examples of each.
3.3.4 Describe what virtual memory is, how it is created and used, and why it is necessary.
3.3.5 Explain what cloud storage is.
3.3.6 Explain the advantages and disadvantages of storing data in the cloud compared with storing it locally.
3.4.1 Explain that a computer needs a network interface card to access a network.
3.4.2 Explain what a MAC address is, what it is for, and how it is structured.
3.4.3a Explain what an IP address is and what it is for.
3.4.3b Explain that there are different types of IP address, including static and dynamic allocation and the characteristics of and differences between IPv4 and IPv6.
3.4.4 Describe the role of a router in a network.

Key terms in Hardware

Magnetic Storage: Magnetic storage is secondary storage in which data is held as the magnetic polarity of tiny areas on a spinning platter or a tape, read and written by a head that moves across the surface as it turns.
Instruction Set: An instruction set is the complete list of machine-code operations a particular processor is built to carry out, defining every command that processor can decode and execute and so making machine code specific to that processor's design.
System Buses: System buses are the sets of parallel wires that carry information between the CPU, memory and input/output units, comprising the address bus carrying memory addresses one way, the data bus carrying data and instructions both ways, and the control bus carrying timing and command signals.
RAM: Random access memory: volatile primary storage, directly accessed by the CPU, which holds the programs and data currently in use and can be both read from and written to; its contents are lost when power is removed.
Cloud Storage: Storage in which data is held on physical servers owned and maintained by a provider and reached remotely over a network or internet connection, rather than on storage attached to the user's own device.
CPU Performance Factors: CPU performance factors are the three properties that decide how quickly a processor completes work — clock speed, the number of cores and the size of the cache — each of which raises performance only under conditions that must be stated.
Secondary Storage: Storage that is not directly accessed by the CPU and is used for the more permanent storage of data; it is generally non-volatile, so data is retained when the power is off, and data needed for processing must first be transferred into RAM.
Memory Data Register: The CPU register that temporarily stores the data or the instruction that has been transferred to or from memory, communicating over the data bus; it is the CPU's doorway for values, not for addresses.
Fetch-Decode-Execute Cycle: The repeating three-stage process by which a CPU runs a program: the next instruction is fetched from memory into the CPU using the program counter, memory address register, address bus, control bus, data bus, memory data register and current instruction register; the control unit then decodes it; and finally the control unit issues the signals that execute it, using the ALU and accumulator where an operation is required.
Arithmetic Logic Unit: The unit within the CPU that performs arithmetic calculations such as addition and subtraction, logical operations such as AND, OR and NOT, and comparisons; its intermediate and final results are commonly placed in the accumulator.
Output Device: Hardware used by a computer to communicate information to a user or to cause a physical action, converting processed data into a form outside the computer such as an image on a screen, a printed page, sound, a solid object or movement.
Solid-State Storage: Solid-state storage is secondary storage with no moving parts, in which data is held by trapping electrons in floating gate transistors using NAND or NOR flash memory, so that the stored charge survives when power is removed.
Control Unit: The unit within the CPU that coordinates the CPU's operations: it decodes each instruction, sends out the control signals that make other components act, and manages the movement of data and instructions around the CPU and to and from memory, so directing the whole fetch-decode-execute cycle.
IP Address: An internet protocol address: an address allocated by the network that identifies a device or connection on that network and allows data to be routed to it; it may be allocated statically, staying the same, or dynamically, so that it can change.
Central Processing Unit: The component of a computer that processes the instructions and data input into the computer so that a result can be output; it controls the computer's operations and contains the control unit, the arithmetic logic unit and a set of registers.
MAC Address: A media access control address: an identifier given to a network interface card at the point of manufacture, usually written in hexadecimal, and made up of a manufacturer code together with a serial code that identifies the individual card.
Von Neumann Architecture: A computer architecture in which program instructions and the data they operate on are held together in the same memory and are accessed using the same addresses, and in which the CPU, memory and input/output components communicate over buses; instructions are fetched and executed one after another in address order unless the flow of control changes.
ROM: Read only memory: non-volatile primary storage that holds instructions or data which must remain when the power is off, typically the instructions a computer needs to start up; in the syllabus model it is read from during normal operation rather than used as working memory.
Current Instruction Register: The CPU register that stores the instruction currently being decoded and executed, so the control unit can read it and work out which operation is required and on what.
Accumulator: The CPU register that stores the intermediate and final results of arithmetic and logic operations, communicating with the arithmetic logic unit; it is one specific named register, not a general term for any register.
Virtual Memory: An area of secondary storage used as though it were RAM when there is not enough RAM for the programs and data currently in use; pages of data are transferred between RAM and virtual memory as they are needed, which allows programs to continue running, but secondary storage is slower than RAM so excessive transfer reduces performance.
Memory Address Register: The CPU register that stores the address of the memory location currently being accessed, and supplies that address to memory along the address bus; it never holds the data found at that address.
Embedded System: A computer system built into a larger device to perform a dedicated function or a limited set of functions; it typically uses a microprocessor or microcontroller running software written for that specific purpose, and differs from a general-purpose computer, which is used to perform many different functions.
Program Counter: The CPU register that stores the address of the next instruction to be fetched; it is updated after each fetch so execution can continue in order, and is loaded with a different address when a jump or branch changes the flow of control.
Input Device: Hardware used to enter data or control signals into a computer system, converting something physical such as a key press, a printed code, an image, a sound or a measured quantity into data the computer can process.
Network Interface Card: The hardware a computer needs in order to access a network; it provides the wired or wireless connection to the network and is given a MAC address at the point of manufacture.
Optical Storage: Optical storage is secondary storage in which data is held as pits and lands along a spiral track on a disc, read by a laser that reflects differently from the two surfaces.
IPv4 and IPv6: IPv4 and IPv6 are the two versions of the Internet Protocol address, IPv4 using 32 bits written as four denary numbers separated by full stops and IPv6 using 128 bits written as eight groups of hexadecimal digits separated by colons.

Common mistakes to avoid

1. “The CPU is the entire computer.” Why wrong A CPU on its own cannot receive a key press, keep a file after a power cut, or show a result. It is one component among several. Correct model The CPU processes instructions and data and controls the computer's operations. The computer is the whole system: CPU, memory, input devices, output devices and secondary storage. Exam-safe “The CPU processes the instructions and data input into the computer so that a result can be output, and controls the computer's operations.” Check: name three things a computer needs that are not the CPU.Any three of: memory (RAM and ROM), input devices, output devices, secondary storage, buses.
2. “A microprocessor is any electronic chip.” Why wrong Memory chips, sound chips and graphics chips are all integrated circuits, and none of them is a microprocessor. Correct model A microprocessor is a type of integrated circuit — one that contains the processing functions of a CPU on a single chip. Every microprocessor is an integrated circuit; not every integrated circuit is a microprocessor. Exam-safe “A microprocessor is a type of integrated circuit on a single chip that contains the processing functions of a CPU.” Check: is a RAM chip a microprocessor?No. It is an integrated circuit, but it stores data rather than containing the processing functions of a CPU.
3. “The MAR stores the instruction.” Why wrong The clue is in the name: memory address register. It holds a where, never a what. Correct model The MAR holds the address of the memory location currently being accessed, and supplies it to memory on the address bus. The instruction that comes back arrives in the MDR. Exam-safe “The MAR stores the address of the memory location currently being accessed.” Check: during the fetch of the instruction at address 100, what is in the MAR?100 — the address, not the instruction stored there.
4. “The MDR stores only memory addresses.” Why wrong It is the exact opposite of the MDR's job, and it leaves nothing to carry values between the CPU and memory. Correct model The MDR temporarily stores the data or instruction transferred to or from memory, over the data bus. In its normal role it does not hold an address. Exam-safe “The MDR temporarily stores the data or instruction that has been transferred to or from memory, using the data bus.” Check: which register is at the CPU end of the data bus?The MDR. The MAR is at the CPU end of the address bus.
5. “The PC stores the current instruction.” Why wrong It confuses two registers and breaks the cycle: if the PC held the current instruction, nothing would know where to go next. Correct model The PC holds the address of the next instruction to be fetched. The current instruction is in the CIR. Exam-safe “The PC stores the address of the next instruction to be fetched, and is updated after each fetch.” Check: two things are wrong in “the PC stores the current instruction”. What are they?It should be an address, not an instruction; and it should be the next one, not the current one.
6. “The CIR stores the address of the next instruction.” Why wrong This is misconception 5 in reverse — the two registers have been swapped. Correct model The CIR holds the current instruction, so the control unit can decode it and then execute it. The address of the next instruction is in the PC. Exam-safe “The CIR stores the instruction currently being decoded and executed.” Check: which register does the control unit read in order to decode?The CIR.
7. “The instruction travels on the address bus.” Why wrong It confuses the address of the instruction with the instruction itself. The address bus carries only addresses. Correct model The address goes out on the address bus; the instruction stored at that address comes back on the data bus. Instructions are stored in memory just like data, so they travel on the data bus. Exam-safe “The address is carried to memory on the address bus, and the instruction is returned on the data bus.” Check: the instruction ADD 301 is being fetched. Which bus carries it, and which bus carried the 301 inside it — and when?The whole instruction travels on the data bus during the fetch. The 301 only reaches the address bus later, during execute, after the instruction has been decoded and 301 has been placed in the MAR.
8. “The ALU controls every CPU component.” Why wrong The ALU is passive: it acts on values put in front of it. Nothing in its name or its definition involves control. Correct model The control unit coordinates the CPU: it decodes instructions, sends control signals and manages the movement of data. The ALU performs arithmetic, logic and comparisons. Exam-safe “The control unit coordinates the operations of the CPU; the ALU performs calculations, logical operations and comparisons.” Check: which unit sends the memory-read signal during a fetch?The control unit, along the control bus.
9. “Execute always means performing arithmetic.” Why wrong Many instructions never touch the ALU — storing a value, loading a value, jumping to an address, sending data to a device. Correct model In execute, the control unit sends the signals needed to carry out whatever the instruction was, using the ALU only when a calculation, logical operation or comparison is required. Exam-safe “The control unit sends the control signals needed to carry out the instruction, using the ALU where a calculation, logical operation or comparison is required.” Check: is the ALU used when executing STO 302?No. Storing the accumulator's value at an address moves data; no calculation is performed.
10. “Higher clock speed guarantees a faster computer.” Why wrong It treats one specification as the whole answer. A CPU waiting for data from RAM does not get more work done by waiting faster. Correct model A higher clock speed can allow more processing operations per second, but overall performance also depends on the number of cores, the size of the cache and what the software is doing. Higher clock speed also increases power consumption and heat. Exam-safe “A higher clock speed means more clock cycles per second, so more instructions can be processed per second — but performance also depends on the cores, the cache and the software.” Check: give one drawback of increasing clock speed.It increases power consumption and generates more heat, which may require extra cooling and shortens battery life.
11. “Twice as many cores always means twice the performance.” Why wrong A second core only helps if there is a second stream of instructions for it to run. Many tasks must be done strictly in order. Correct model More cores allow more than one instruction stream to be processed at the same time, and performance improves when the software can divide the work effectively. Adding cores does not guarantee a proportional improvement. Exam-safe “More cores can process more instruction streams at once, but only if the software can divide its work between them.” Check: a program is single-threaded. How much faster will it run on eight cores than on one?Essentially no faster. One instruction stream can only be processed by one core, so the other seven have nothing to do.
12. “Cache and RAM are identical.” Why wrong If they were the same there would be no reason to have both, and no reason cache improves performance. Correct model Cache is small, fast memory inside or very close to the CPU that holds frequently or recently used data and instructions. RAM is much larger, slower and further away, and holds everything currently in use. Cache exists to reduce how often the slower RAM has to be accessed. Exam-safe “Cache is small, fast memory close to the CPU holding frequently used data and instructions, so fewer accesses to the slower RAM are needed.” Check: state two differences between cache and RAM.Any two of: cache is smaller; cache is faster to access; cache is inside or closer to the CPU; cache holds only frequently or recently used data and instructions; cache costs more per unit of storage.
13. “An instruction set is one program.” Why wrong It confuses the vocabulary with a sentence written in it. Every program on a CPU draws from the same set. Correct model An instruction set is the complete list of all the commands a CPU can process, written in machine code. It belongs to the processor design, not to any one program. Exam-safe “An instruction set is a list of all the commands that can be processed by a CPU, and the commands are machine code.” Check: two programs run on the same CPU. Do they use the same instruction set?Yes. Both must be built from the instructions that particular CPU can process, even though the two programs do completely different things.
14. “An embedded system has no software.” Why wrong A washing machine controller executes millions of instructions to run a wash cycle. Something has to tell it what to do. Correct model An embedded system runs software written for its specific purpose. What makes it different is that the user does not choose, install or replace that software — not that there is none. Exam-safe “An embedded system performs a dedicated function using software designed for that specific purpose.” Check: what does “dedicated” actually describe — the number of instructions, or the purpose?The purpose. The system exists to do one job; it still executes very many instructions to do it.
15. “A touch screen is only an output device.” Why wrong It counts the display and ignores the touch sensing, which is how data and control signals get in. Correct model A touch screen is both. The display communicates information, which is output; the touch sensing enters data and control signals, which is input. Exam-safe “A touch screen is both an input and an output device: it displays information and it accepts input when it is touched.” Check: name the three touch-screen technologies on the syllabus.Resistive, capacitive and infra-red.
16. “A sensor makes the final control decision.” Why wrong A sensor has no stored target value and nothing to compare against. It cannot know whether a reading is acceptable. Correct model A sensor measures a physical property and produces data. A microprocessor compares that data with a stored value and decides. If a condition is met, a signal is sent to an actuator or another output device. Exam-safe “The sensor sends readings to the microprocessor, which compares them with a stored value and, if necessary, sends a signal to an actuator.” Check: fill the gaps — sensor → ______ → ______.sensor → microprocessor (compares with a stored value) → actuator (or another output device).
17. “An actuator is an input device.” Why wrong It has the direction backwards. An actuator receives a signal; it does not produce data for the computer. Correct model An actuator is an output device that converts a control signal from the computer into a physical action — movement, opening or closing, switching on or off. Exam-safe “An actuator is an output device that converts a signal from the microprocessor into a physical action.” Check: in a greenhouse system, which is the sensor and which is the actuator — the device that measures the temperature, or the motor that opens the vent?The device that measures the temperature is the sensor (input); the motor that opens the vent is the actuator (output).
18. “RAM is permanent storage.” Why wrong RAM is volatile. Anything in it disappears the instant the power is removed — which is why an unsaved document is lost in a power cut. Correct model RAM holds the programs and data currently in use and can be read from and written to, but it is volatile. Permanent storage is the job of secondary storage. Exam-safe “RAM is volatile primary storage holding the programs and data currently in use; its contents are lost when power is removed.” Check: a file is edited but not saved, and the power fails. Where was the file, and why is it gone?It was in RAM, which is volatile, so its contents were lost. It had not yet been written to secondary storage.
19. “ROM is ordinary working memory.” Why wrong Working memory has to be written to constantly as programs run. That is not what ROM is for in this model. Correct model ROM is non-volatile primary storage holding instructions and data that must always be present, such as the start-up instructions. It is read from during normal operation; RAM is the working memory. Exam-safe “ROM is non-volatile and stores the instructions the computer needs at start-up; it is read from during normal operation rather than used as working memory.” Check: which of RAM and ROM is usually much larger, and why does that make sense?RAM, because it has to hold every program and file currently in use, whereas ROM only has to hold a fixed set of start-up instructions.
20. “An SSD has no transistors because it has no moving parts.” Why wrong It confuses “no mechanical movement” with “no components”. Solid-state storage is built almost entirely from transistors. Correct model Solid-state storage uses NAND or NOR flash technology in which transistors act as control gates and floating gates. “No moving parts” means nothing mechanical moves — electrons still do. Exam-safe “Solid-state storage has no moving parts and uses NAND or NOR flash technology, with transistors acting as control gates and floating gates.” Check: what keeps the electrons on the floating gate when the power is off?The insulating layers surrounding it. That is why solid-state storage is non-volatile.
21. “Virtual memory is extra RAM.” Why wrong No physical memory has been added. Virtual memory is space on a drive being used as if it were RAM. Correct model Virtual memory is an area of secondary storage used when there is not enough RAM. Pages are transferred between RAM and virtual memory as they are needed, and the CPU still only works with RAM. Exam-safe “Virtual memory uses part of secondary storage when there is insufficient RAM, with pages of data transferred between the two as they are needed.” Check: where does virtual memory physically exist?In secondary storage — on the hard disk drive or solid-state drive.
22. “Virtual memory always improves performance.” Why wrong Every page transfer costs time, because secondary storage is slower than RAM. Using virtual memory heavily makes a machine slower, not faster. Correct model Virtual memory lets programs continue running when RAM is full, instead of stopping. The correct comparison is “slow but still working” against “unable to continue” — never against “faster”. Exam-safe “Virtual memory allows programs to continue running when RAM is insufficient, but because secondary storage is slower than RAM, frequent page transfers reduce performance.” Check: a computer is slow because it is using virtual memory heavily. Does a bigger hard disk fix it?No. More RAM is needed, so that fewer pages have to be transferred at all. A bigger disk only makes the slow store larger.
23. “Cloud storage has no physical hardware.” Why wrong The word “cloud” describes where the data is from the user's point of view, not what it is made of. Correct model Cloud storage keeps data on physical servers in a data centre, owned and maintained by a provider, and reached over a network or internet connection. Exam-safe “Cloud storage stores data on remote physical servers, which are accessed over a network or internet connection.” Check: give one disadvantage of cloud storage that follows directly from it being remote.It cannot be reached without a network connection; transferring very large amounts of data can also be slow.
24. “MAC and IP addresses are interchangeable.” Why wrong They identify different things, are given by different parties, and are used for different jobs. A network needs both. Correct model A MAC address identifies the network interface card and is given at manufacture. An IP address identifies a device on a network, is allocated by the network, and is what data is routed to. Exam-safe “A MAC address identifies the network interface card and is assigned at manufacture, whereas an IP address is allocated by the network and allows data to be routed to the device.” Check: which of the two does a router read in order to forward a packet?The destination IP address in the packet's header.
25. “A MAC address changes on every network.” Why wrong It describes the IP address. The MAC address belongs to the card, not to the network. Correct model The MAC address is given at the point of manufacture and does not normally change when the device joins a different network. It is the IP address that is allocated afresh by each network. Exam-safe “A MAC address is assigned at manufacture and stays with the network interface card; the device's IP address is allocated by whichever network it joins.” Check: a laptop is used at school and then at home. Which address changes?The IP address. The MAC address stays the same.
26. “IPv6 is simply faster IPv4.” Why wrong Speed depends on the connection, the hardware and the route taken — not on how many bits an address contains. Correct model IPv6 uses a 128-bit address written as hexadecimal groups, against IPv4's 32-bit address written as four denary groups. The difference that matters is the size of the address space. Exam-safe “An IPv6 address is 128 bits, compared with 32 bits for IPv4, so IPv6 provides a much larger number of possible addresses.” Check: state the two differences worth writing about IPv4 and IPv6.The length of the address (32 bits against 128 bits) and how it is written (four denary groups against hexadecimal groups). The consequence is a much larger address space.
27. “A router and a NIC are the same device.” Why wrong One is inside the device and gives it a connection; the other is separate hardware that joins networks together. Correct model A NIC is the hardware a computer needs to access a network, and carries the MAC address. A router reads the destination address in each packet, selects a route and forwards it; it can also assign IP addresses and connect a local network to the internet. Exam-safe “A NIC gives one device its connection to a network; a router forwards packets between networks towards their destination.” Check: which of the two provides the MAC address, and which can provide an IP address?Neither provides the MAC address — the manufacturer does, and it is stored on the NIC. The router can allocate an IP address on its own network.

Examiner tips

Read the command word in the outcome. The syllabus itself distinguishes understand and state outcomes from describe and explain outcomes. Outcome 3.1.2b says describe the process — that means an ordered sequence with component names. Outcome 3.1.3 says explain how they can affect performance — that means a causal chain, not a list of three words. Outcome 3.3.6 says in comparison to — that means both sides of the comparison must appear in your answer.
Calculators are not permitted, so nothing in Topic 3 needs one. Where a number appears — 32 bits, 128 bits, 48 bits, a clock speed in GHz — you are being asked to compare or to interpret, never to compute. If you find yourself trying to work out \(2^{128}\), you have misread the question.
About step 8 — where the PC is increased. Different textbooks place the increment at slightly different points within the fetch stage: some show it immediately after the address is copied into the MAR, others (as here) show it once the instruction has safely reached the CIR. Either is acceptable, because the effect is identical — by the end of the fetch stage the PC holds the address of the next instruction. What is not acceptable is being inconsistent inside one answer, or leaving the increment out altogether. This chapter, including the interactive stepper and the register tracker, uses one position throughout: the PC is increased at the end of the fetch stage.
Avoid absolute claims. “An SSD never fails” and “optical discs last for ever” are both wrong, and both are easy to write by accident when you are trying to praise a technology. Solid-state storage has a finite number of write cycles per cell and can fail; optical discs scratch, warp and degrade. Write “more resistant to physical shock”, not “indestructible”.
Apply, do not list. A comparison question is marked on whether each point is applied to the scenario in front of you. “Cloud storage can be accessed remotely” is a fact; “the photographer travels between locations, so cloud storage lets them reach the files from a laptop on site rather than only from the office computer” is an answer. Take one detail from the scenario into every point you make.

Frequently asked questions

What is the difference between RAM and ROM?

RAM (random access memory) is volatile primary storage that holds the programs and data currently in use, and can be both read from and written to; its contents are lost when power is removed. ROM (read only memory) is non-volatile primary storage that holds instructions, such as the start-up instructions, which must remain when the power is off, and is read from during normal operation. A computer needs both: without RAM there is nowhere to work, and without ROM there is nothing to tell the machine what to do when it is first switched on.

What is the difference between the PC, MAR, MDR and CIR registers?

The program counter (PC) holds the address of the next instruction to be fetched. The memory address register (MAR) holds the address of the memory location currently being accessed, supplying it to memory on the address bus. The memory data register (MDR) temporarily stores the data or instruction transferred to or from memory over the data bus. The current instruction register (CIR) holds the instruction currently being decoded and executed. PC and MAR hold a "where"; MDR and CIR hold a "what".

Why does a higher clock speed not always mean a faster computer?

A higher clock speed can allow more processing operations per second, but overall performance also depends on the number of cores, the size of the cache, and what the software is actually doing. A CPU with many cores only benefits from them if the software can divide its work between them, and cache reduces how often the CPU has to wait for the slower RAM. Higher clock speed also increases power consumption and heat, so it is only one of several factors.

What is the difference between magnetic, optical and solid-state storage?

Magnetic storage, such as a hard disk drive, uses electromagnets in a read/write head to magnetise regions of a spinning platter in one of two directions, representing 1s and 0s. Optical storage, such as a CD, DVD or Blu-ray, uses a laser to read pits and lands on a disc's surface, which reflect light differently. Solid-state storage, such as an SSD, SD card or USB flash drive, has no moving parts and stores data by trapping electrons on floating gate transistors using NAND or NOR flash technology.

What is the difference between a MAC address and an IP address?

A MAC address identifies the network interface card (NIC) and is assigned at manufacture, usually written in hexadecimal; it does not normally change when a device joins a different network. An IP address identifies a device or connection on a particular network and is allocated by that network, either statically or dynamically, so it can change from network to network. A router reads the destination IP address in a packet's header to decide how to forward it.

Why is virtual memory not simply extra RAM?

Virtual memory is an area of secondary storage used as though it were RAM when there is not enough physical RAM for the programs and data in use; pages of data are transferred between RAM and virtual memory as they are needed. No physical memory has actually been added, and because secondary storage is much slower than RAM, frequent page transfers reduce performance rather than improving it. Virtual memory lets programs keep running when RAM is full instead of stopping, but it does not make the computer faster.

What is an embedded system, and does it run without software?

An embedded system is a computer system built into a larger device to perform a dedicated function or a limited set of functions, typically using a microprocessor running software written for that specific purpose. It is not true that it runs without software — a device such as a washing machine controller executes very many instructions to run a wash cycle. What makes it "dedicated" is its fixed purpose, and that the user does not choose, install or replace that software, not the absence of software itself.

Syllabus reference and sources

Written against: Cambridge O Level Computer Science (2210) 2026–2028 Syllabus (Subject Content, Topic 3: Hardware).

Written by: Academiq Edu Instructor Panel

Source documents

All educational content, structured explanations, diagrams, worked examples, and pedagogical materials contained within this chapter revision note are the exclusive intellectual property of Academiq Edu. Unauthorized reproduction, distribution, resale, or extraction of this content without prior written permission is strictly prohibited under international copyright laws. Cambridge Assessment International Education (CAIE) is a registered trademark of Cambridge University Press & Assessment. This revision guide is independently authored by the Academiq Edu Instructor Panel for educational purposes and is not affiliated with or endorsed by Cambridge Assessment International Education.

Verified Curriculum Content

Every chapter note, MCQ explanation, and structured mark scheme is rigorously vetted by Cambridge curriculum specialists.