Hardware: fetch-execute cycle and the effect of CPU characteristics on performance (clock speed, cores, cache size)

What you'll learn

This revision guide covers how the Central Processing Unit (CPU) executes instructions and the factors that determine processor performance. You'll understand the fetch-execute cycle in detail and learn how clock speed, number of cores, and cache size affect how quickly a computer can process data—all essential knowledge for CIE IGCSE Computer Science exam questions on hardware.

Key terms and definitions

Central Processing Unit (CPU) — the hardware component that processes instructions and data, often called the "brain" of the computer

Fetch-Execute Cycle — the sequence of operations performed by the CPU to execute a single program instruction, repeated continuously while a program runs

Clock Speed — the number of fetch-execute cycles a CPU can perform per second, measured in hertz (Hz)

Core — an independent processing unit within a CPU that can execute its own fetch-execute cycle simultaneously with other cores

Cache — very fast memory located on or near the CPU chip that stores frequently used instructions and data for quick access

Control Unit (CU) — the part of the CPU that manages the fetch-execute cycle and coordinates all processor activities

Arithmetic Logic Unit (ALU) — the part of the CPU that performs arithmetic calculations and logical operations

Register — extremely fast storage location inside the CPU used to temporarily hold data, addresses, and instructions during processing

Core concepts

The fetch-execute cycle

The fetch-execute cycle is the fundamental process by which a CPU executes program instructions. Every instruction that runs on a computer goes through this cycle. The cycle has distinct stages that occur in sequence:

Stage 1: Fetch

• The Control Unit retrieves the next instruction from main memory (RAM)
• The address of the instruction is held in the Program Counter (PC) register
• The instruction is copied from RAM to the Current Instruction Register (CIR)
• The Program Counter increments to point to the next instruction

Stage 2: Decode

• The Control Unit interprets the instruction in the CIR
• It determines what operation needs to be performed
• It identifies which data or memory addresses are needed

Stage 3: Execute

• The instruction is carried out
• If it's a calculation, the Arithmetic Logic Unit performs it
• If data needs moving, the Control Unit manages the transfer
• Results are stored in registers or written back to memory

The cycle then repeats, fetching the next instruction. This happens billions of times per second in modern processors.

The role of registers in the fetch-execute cycle

Registers are the fastest type of storage in a computer, located inside the CPU itself. Key registers involved in the fetch-execute cycle include:

Program Counter (PC)

• Holds the memory address of the next instruction to be fetched
• Automatically increments after each fetch
• Can be modified by jump or branch instructions

Current Instruction Register (CIR)

• Stores the instruction currently being decoded and executed
• Holds the instruction throughout the decode and execute stages

Memory Address Register (MAR)

• Holds the address of the memory location being accessed
• Used when reading from or writing to RAM

Memory Data Register (MDR)

• Temporarily stores data being transferred to or from memory
• Acts as a buffer between the CPU and main memory

Accumulator

• Stores the results of calculations performed by the ALU
• Holds intermediate values during arithmetic and logical operations

Clock speed and CPU performance

Clock speed measures how many fetch-execute cycles a processor can complete per second. It's measured in:

• Hertz (Hz) — one cycle per second
• Megahertz (MHz) — one million cycles per second
• Gigahertz (GHz) — one billion cycles per second

A modern CPU might have a clock speed of 3.5 GHz, meaning it can complete 3.5 billion fetch-execute cycles per second.

Impact on performance:

• Higher clock speed = more instructions processed per second
• A CPU running at 3 GHz can theoretically execute twice as many instructions per second as a 1.5 GHz CPU
• Programs run faster when clock speed increases
• Each "tick" of the system clock triggers the next stage in the fetch-execute cycle

Limitations:

• Clock speed cannot increase indefinitely due to heat generation
• Beyond approximately 4-5 GHz, cooling becomes impractical for consumer devices
• Power consumption increases dramatically with higher clock speeds
• Comparing clock speeds only makes sense between processors of the same family or architecture

Number of cores and parallel processing

A core is an independent processing unit within a CPU. Modern processors contain multiple cores, each capable of running its own fetch-execute cycle simultaneously.

Single-core processors:

• Execute one instruction stream at a time
• Can only run one program thread simultaneously
• Simpler design, less expensive to manufacture

Multi-core processors:

• Dual-core: 2 independent cores
• Quad-core: 4 independent cores
• Octa-core: 8 independent cores
• Higher-end processors may have 16, 32, or more cores

Impact on performance:

• Multiple cores enable parallel processing—executing multiple instruction streams simultaneously
• A quad-core processor can theoretically execute four times as many instructions per second as a single-core processor at the same clock speed
• Different programs can run on different cores, improving multitasking
• Single programs can be split across cores if written to support parallel processing

Limitations:

• Not all software is designed to use multiple cores effectively
• Some programs can only run on a single core, gaining no benefit from additional cores
• Coordinating work between cores adds overhead
• Doubling the cores doesn't double overall system performance—typical improvement is 50-80%

Cache size and memory access speed

Cache is extremely fast memory located on or very close to the CPU chip. It stores copies of frequently used data and instructions to reduce the time spent accessing main memory (RAM).

Cache levels:

• L1 cache — smallest (typically 32-64 KB per core), fastest, built into each core
• L2 cache — medium size (typically 256 KB-1 MB per core), very fast, may be dedicated to each core
• L3 cache — largest (typically 4-32 MB), fast, shared between all cores

How cache works:

• When the CPU needs data, it checks L1 cache first
• If not found (cache miss), it checks L2, then L3
• If still not found, it must fetch from main RAM (much slower)
• When data is retrieved from RAM, a copy is stored in cache for potential reuse
• Cache stores both instructions and data that are likely to be needed again soon

Impact on performance:

• Cache access is typically 10-100 times faster than RAM access
• Larger cache means more data can be stored close to the CPU
• Higher cache hit rate (finding data in cache rather than RAM) significantly improves performance
• Programs with good locality of reference (repeatedly using the same data) benefit most from cache

Limitations:

• Cache is expensive to manufacture, limiting how much can be included
• Diminishing returns—doubling cache size doesn't double performance
• Cache management adds complexity to processor design
• Some programs don't benefit significantly from larger cache if they constantly access different memory locations

Comparing CPU performance

When comparing processors, you must consider all three factors together:

Scenario 1: Processor A—3.0 GHz, 4 cores, 8 MB cache Scenario 2: Processor B—3.5 GHz, 2 cores, 4 MB cache

• Processor A has more cores, making it better for multitasking and parallel applications
• Processor B has higher clock speed, potentially faster for single-threaded programs
• Processor A's larger cache may provide better performance for programs that reuse data frequently
• The "better" processor depends entirely on the intended use case

Other factors affecting performance:

• CPU architecture and design efficiency
• Manufacturing process technology (e.g., 7nm vs 14nm)
• Instruction set and features
• System bottlenecks (slow RAM, storage, or graphics)

Worked examples

Example 1: Fetch-execute cycle sequence

Question: Describe the three stages of the fetch-execute cycle. [6 marks]

Mark scheme answer:

Fetch stage:

• The address in the Program Counter is copied to the Memory Address Register [1]
• The instruction at that address is copied from memory to the Memory Data Register [1]
• The instruction is then copied to the Current Instruction Register / the Program Counter is incremented [1]

Decode stage:

• The Control Unit interprets/decodes the instruction [1]
• It determines what operation is required and what data is needed [1]

Execute stage:

• The instruction is carried out [1]
• This might involve the ALU performing a calculation or data being moved between locations [1]

(Any 6 marks from the points above)

Example 2: CPU characteristics and performance

Question: A computer manufacturer offers two laptop models with different processors:

• Model X: 2.8 GHz dual-core processor with 4 MB cache
• Model Y: 2.4 GHz quad-core processor with 8 MB cache

A student wants to run video editing software that can use multiple cores simultaneously.

(a) Explain which processor would be more suitable for this task. [3 marks]

(b) State two reasons why clock speed alone is not sufficient for comparing CPU performance. [2 marks]

Mark scheme answer:

(a)

• Model Y would be more suitable [1]
• The video editing software can use multiple cores / supports parallel processing [1]
• Model Y has four cores compared to two cores in Model X, allowing more simultaneous processing [1]
• The larger cache (8 MB) will also help with the large video files being processed [1]

(Any 3 marks)

(b)

• The number of cores affects how many instructions can be processed simultaneously [1]
• Cache size affects how quickly data can be accessed [1]
• Different CPU architectures have different efficiency [1]
• Not all programs can utilise higher clock speeds effectively [1]

(Any 2 marks)

Example 3: Cache memory

Question: Explain why cache memory improves CPU performance. [4 marks]

Mark scheme answer:

• Cache stores copies of frequently used data and instructions [1]
• Cache is located on or very close to the CPU / integrated into the processor chip [1]
• Accessing cache is much faster than accessing main memory/RAM [1]
• This reduces the time the CPU spends waiting for data [1]
• The CPU checks cache first before accessing main memory [1]
• More instructions can be completed per second [1]

(Any 4 marks)

Common mistakes and how to avoid them

• Confusing the order of fetch-execute cycle stages — Remember the acronym FDE: Fetch, Decode, Execute. The sequence never changes and each stage must complete before the next begins.

• Thinking more cores always means better performance — More cores only help if software is designed to use them. A single-threaded program runs no faster on an 8-core processor than a single-core processor at the same clock speed. Always consider the application when comparing processors.

• Stating that "cache stores frequently used programs" — Cache stores frequently used data and instructions, not entire programs. Be precise: cache holds copies of small portions of programs and data that are actively being processed.

• Believing higher clock speed always means better performance — A 3 GHz quad-core processor will outperform a 4 GHz single-core processor for multitasking. Consider all factors: clock speed, cores, cache, and architecture together.

• Mixing up register types and their functions — The Program Counter holds the address of the next instruction, while the Current Instruction Register holds the actual instruction being executed. Memory Address Register holds addresses; Memory Data Register holds data. Learn the specific role of each register.

• Assuming cache makes all programs faster equally — Programs that repeatedly access the same data benefit greatly from cache. Programs that constantly access new data locations see minimal benefit. Cache effectiveness depends on locality of reference in the program's design.

Exam technique for "Hardware: fetch-execute cycle and the effect of CPU characteristics on performance (clock speed, cores, cache size)"

• "Describe" questions about the fetch-execute cycle — Give a clear sequence of events with specific details. Mention the relevant registers (PC, CIR, MAR, MDR) and the components involved (Control Unit, ALU). Aim for 2-3 points per stage if 6 marks are available.

• "Explain why" questions about CPU characteristics — Always link the characteristic to its effect on performance. Don't just state "more cache is faster"—explain that it reduces memory access time, allowing more instructions to complete per second. Use cause-and-effect structure in your answers.

• Comparison questions — When comparing processors, systematically consider each characteristic (clock speed, cores, cache) and explain which processor has the advantage for that factor. Then provide an overall judgement based on the specific use case given in the question.

• Command word precision — "State" requires a single fact without explanation (1 mark). "Describe" requires characteristics or steps (2-3 marks). "Explain" requires reasons linked to outcomes (2-4 marks). Match your answer length and detail level to the command word and mark allocation.

Quick revision summary

The CPU executes instructions through the fetch-execute cycle: fetching the instruction from memory (using PC and MAR), decoding what operation is required (in the CU), and executing it (using the ALU if needed). CPU performance depends on clock speed (cycles per second), number of cores (parallel processing capability), and cache size (fast data access). Higher values for each characteristic generally improve performance, but the actual benefit depends on the software being run and how efficiently it utilises these features. Consider all factors together when comparing processors.