Skip to topic
    ← Back to course topics

    The processor and its components — AQA A-Level Computer Science

    The central processing unit (CPU) executes program instructions via the Fetch-Decode-Execute cycle using components including the arithmetic logic unit (ALU), control unit (CU), registers, and system buses. Mastery of registers such as the program counter, memory address register, and memory buffer register is essential for explaining instruction flow and processor performance in AQA A-Level Computer Science.

    The processor and its components explained

    The processor, or central processing unit (CPU), is the component that fetches, decodes and executes the instructions of a program.

    Read the full explanation

    It works through the fetch-decode-execute cycle, reading an instruction from memory at the address held in the program counter, decoding it in the control unit, and carrying out the operation using the arithmetic logic unit and registers. Its major components are the control unit, which coordinates activity and sends control signals; the arithmetic logic unit, which performs calculations and logical comparisons; and registers, small high-speed storage locations such as the program counter, accumulator and memory address register that hold data and addresses during processing. A clock controls the timing of these operations, and the cache provides fast temporary storage of frequently used data.

    Your focus

    1. Explain the role and operation of a processor and its major components:

    Which command word is this?

    Set out the causes, reasons, or mechanics of a component. For instance, explaining the role of the CIR requires stating what it holds, how the instruction is partitioned into opcode and operand, and which unit decodes it.

    The processor and its components exam tips

    Topic Overview

    The processor is the central computational engine of a computer system, responsible for fetching, decoding, and executing program instructions stored in main memory. In AQA A-Level Computer Science, this topic delves into the internal Von Neumann architecture, detailing how the Arithmetic Logic Unit (ALU), Control Unit (CU), internal clock, and dedicated registers work harmoniously to process data at billions of cycles per second.

    Understanding this microarchitecture explains the physical bottlenecks of computing and bridges the gap between high-level code and low-level hardware execution. Mastery of system buses, register transfer notation, and performance factors equips students to analyse system design trade-offs and evaluate how modern processors balance clock speed, cache hierarchies, and multi-core parallelism.

    Key Concepts
    • →The Von Neumann architecture features shared memory for both data and instructions, communicating with the CPU via shared system buses.
    • →Dedicated registers hold temporary data and instructions: PC holds the address of the next instruction, MAR holds memory addresses, MBR holds data/instructions read from or written to memory, CIR holds the instruction being decoded, and ACC accumulates ALU calculation results.
    • →System buses comprise the unidirectional Address Bus, bidirectional Data Bus, and bidirectional/individual line Control Bus carrying signals like memory read, memory write, bus request, and clock pulses.
    • →Processor performance is governed by clock frequency, word length, bus widths, cache size and levels (L1, L2, L3), and number of processing cores alongside instruction pipelining.
    Marking Points
    • Describes the processor as fetching, decoding and executing instructions
    • Explains the role of the control unit in coordinating and sending control signals
    • Identifies the arithmetic logic unit as performing arithmetic and logical operations
    • Explains that registers temporarily hold data, addresses and instructions during processing
    • Links the components to the fetch-decode-execute cycle
    Examiner Tips
    • 💡Name each major component and state its specific function
    • 💡Link components to the fetch-decode-execute cycle to show how they operate together
    • 💡Use precise terms such as program counter and memory address register rather than vague words like 'storage'
    • 💡Use precise Register Transfer Notation (RTN) with square brackets denoting contents, e.g. MBR <- [[MAR]], when detailing the Fetch-Decode-Execute cycle to secure top marks.
    • 💡When asked how cache affects performance, explicitly reference 'spatial locality' and 'temporal locality' to demonstrate an advanced technical vocabulary.
    • 💡Distinguish clearly between the bus request line and the bus grant line when explaining how control bus lines coordinate direct memory access (DMA) or multi-master systems.
    Common Mistakes
    • Describes the processor as storing programs permanently; memory holds programs and the processor only executes them
    • Confuses the control unit with the arithmetic logic unit; the control unit directs operations and the ALU performs them
    • Omits the link between components and the fetch-decode-execute cycle; explain how each part contributes to a single instruction
    • Thinking the Address Bus is bidirectional: The CPU generates memory addresses to access RAM or I/O controllers, so the address bus transmits signals in one direction only (CPU to memory/controllers), unlike the Data Bus which moves data both ways.
    • Believing cache memory stores entire software applications: Cache is a tiny, fast SRAM store positioned near or on the CPU die that only retains frequently and recently used instructions and data according to locality of reference.
    • Confusing the Current Instruction Register (CIR) with the Memory Buffer Register (MBR): The MBR acts as a general gateway holding any data or instruction fetched from RAM, whereas the CIR exclusively holds the specific instruction currently being split into opcode and operand during decode.
    Revision Plan
    1. 1Day 1-3: Memorise each register's precise name and role, and practice writing the Fetch-Decode-Execute cycle steps repeatedly using correct Register Transfer Notation.
    2. 2Day 4-6: Diagram the three system buses, tracing the direction of signals between the CPU, RAM, and I/O controllers for read and write operations.
    3. 3Day 7-9: Work through calculations involving address bus width, data bus width, and clock speed throughput, ensuring unit conversions (Hz, MHz, GHz, bytes, GiB) are flawless.
    4. 4Day 10-12: Complete 6-mark and 9-mark evaluation questions examining trade-offs between clock speed, cache size, multi-core architecture, and pipelining under timed conditions.
    Exam Question Types
    • 📋Short-answer definitions: Stating the function of a named register (e.g. CIR, MAR, PC) or system bus with 1-2 mark precision.
    • 📋Step-by-step tracing: Describing register states during specific stages of the Fetch-Decode-Execute cycle (often tested via RTN or sequence ordering).
    • 📋Calculations: Determining directly addressable memory given address bus width, or calculating data transfer rates given bus clock frequency and bus width.
    • 📋Extended evaluation essays: Evaluating hardware upgrades (e.g. adding cores vs increasing cache vs increasing clock frequency) for specific use cases like gaming, video editing, or database servers.
    Command Word Expectations (AQA)
    Explain

    Set out the causes, reasons, or mechanics of a component. For instance, explaining the role of the CIR requires stating what it holds, how the instruction is partitioned into opcode and operand, and which unit decodes it.

    Describe

    Provide a factual, detailed step-by-step account of a process without necessarily justifying why. In an FDE cycle question, this means writing down the sequence of register transfers in chronological order.

    Evaluate

    Weigh up advantages, disadvantages, and real-world limitations of processor characteristics (e.g., multicore vs pipelining), leading to a justified, balanced conclusion directly tied to the scenario.

    How Students Lose Marks (Examiner Pitfalls)
    Pitfall: Confusing the roles of the Memory Address Register (MAR) and Memory Buffer Register (MBR/MDR) during the Fetch stage.
    ❌ Weak Answer (Loses Marks):The Program Counter copies the address to the MDR, then the instruction is read from memory into the MAR.
    Example improved answer:The contents of the Program Counter (PC), which holds the memory address of the next instruction to be fetched, are copied to the Memory Address Register (MAR) across the internal bus. This address is placed onto the address bus. The instruction located at that address in main memory is returned across the data bus and copied directly into the Memory Buffer Register (MBR).
    Examiner Tip: Always remember that addresses strictly travel through the MAR via the unidirectional address bus, whereas instructions and data travel through the MBR via the bidirectional data bus.
    Pitfall: Stating that doubling the number of CPU cores will automatically double the execution speed of any program.
    ❌ Weak Answer (Loses Marks):A quad-core processor is always four times faster than a single-core processor because it has four brains doing work.
    Example improved answer:A multi-core processor can execute instructions in parallel only if the software is written to support multi-threading. Performance rarely scales linearly due to inter-core communication overhead, sequential code bottlenecks described by Amdahl's Law, and contention for shared system resources such as the system bus and L3 cache.
    Examiner Tip: In evaluation or 'explain' questions about processor architecture, always state that parallel scaling is constrained by software architecture (parallelisable algorithms) and hardware overhead.
    Step-by-Step Worked Solutions

    Question: Describe the Fetch stage of the Fetch-Decode-Execute cycle using Register Transfer Notation (RTN) and explain the role of the Program Counter (PC).

    1. 1.Step 1: State the initial register transfer where the address of the next instruction is transferred from the PC to the MAR: MAR <- [PC].
    2. 2.Step 2: Show the simultaneous update of the PC to point to the subsequent memory location and the retrieval of data from primary memory: PC <- [PC] + 1; MBR <- [[MAR]].
    3. 3.Step 3: Transfer the fetched instruction from the Memory Buffer Register to the Current Instruction Register for decoding: CIR <- [MBR].
    4. 4.Step 4: Conclude by explaining that the PC acts as a memory pointer, ensuring instructions are fetched in strict sequential order unless altered by jump, branch, or interrupt signals.
    Final Answer: The Fetch stage in RTN is: MAR <- [PC]; PC <- [PC] + 1; MBR <- [[MAR]]; CIR <- [MBR]. The PC maintains sequential instruction sequencing by tracking memory addresses.

    Question: A processor architecture features a 36-bit address bus and a 64-bit data bus. Calculate the maximum directly addressable memory capacity in gibibytes (GiB), assuming memory is byte-addressed, and explain the significance of the data bus width.

    1. 1.Step 1: Calculate the total number of unique addressable memory locations using 2^n, where n is the width of the address bus: 2^36 = 68,719,476,736 locations.
    2. 2.Step 2: Because memory is byte-addressed (1 byte per addressable unit), calculate total capacity in bytes: 2^36 bytes. Convert to gibibytes (GiB) by dividing by 2^30: 2^36 / 2^30 = 2^6 = 64 GiB.
    3. 3.Step 3: State the significance of the 64-bit data bus: it determines the word size that can be transferred in a single read or write cycle between the CPU and memory, allowing 8 bytes of data or instruction code to be fetched simultaneously.
    Final Answer: Maximum directly addressable memory is 64 GiB. The 64-bit data bus allows 8 bytes to be transferred in a single cycle, directly increasing memory throughput.
    Active Recall Memory Test
    What does the Program Counter (PC) hold?
    Key Fact: The memory address of the next instruction to be fetched and executed.
    Why is the address bus unidirectional whereas the data bus is bidirectional?
    Key Fact: The address bus only carries memory addresses generated by the CPU to memory/IO, while the data bus must transmit data both to (write) and from (read) the CPU.
    What two components is an instruction split into inside the Current Instruction Register (CIR)?
    Key Fact: The opcode (operation to be performed) and the operand (data or address to be operated on).
    What condition causes instruction pipelining to stall or flush?
    Key Fact: A conditional branch or jump instruction that alters the Program Counter, rendering speculatively fetched instructions invalid.
    Frequently Asked Questions
    What is the difference between the MBR and the MDR?
    There is no functional difference; they are two names for the exact same register. AQA A-Level mark schemes accept both Memory Buffer Register (MBR) and Memory Data Register (MDR). This register acts as a two-way buffer holding any instruction or data value recently read from, or about to be written to, main memory.
    How does cache memory speed up processor execution?
    Cache memory is constructed from high-speed Static RAM (SRAM) and sits directly on or adjacent to the processor chip. It exploits temporal and spatial locality by holding frequently or recently accessed instructions and data. Because accessing SRAM takes a fraction of the clock cycles required to access off-chip Dynamic RAM (DRAM), CPU idle time during memory fetch cycles is significantly reduced.
    Why does doubling the clock speed not always double system performance?
    While doubling the clock speed increases the rate of basic ALU and register operations, the CPU is frequently delayed by the 'Von Neumann bottleneck'. Main memory (RAM) and secondary storage operate at much lower clock speeds, creating wait states (cache misses) where the CPU must idle. Furthermore, increased clock speed generates exponentially more heat and power consumption, often triggering thermal throttling.
    What is the difference between Harvard and Von Neumann architecture?
    Von Neumann architecture uses a single shared physical memory space and common buses for both instructions and data, which can lead to bus contention known as the Von Neumann bottleneck. In contrast, Harvard architecture features physically separate memory units and dedicated buses for instructions and data, enabling simultaneous instruction fetch and data read/write. Contemporary processors often use a 'Modified Harvard' architecture with separate L1 instruction and data caches backed by unified main memory.