CPU Architecture and the Fetch-Execute Cycle
What Is the CPU?
The CPU is the component that carries out every instruction a program contains. It processes data, performs calculations, makes comparisons and controls the flow of information between all other parts of the system. Without the CPU, no program can run.
Most CPUs follow the design principles of Von Neumann architecture, named after the mathematician John von Neumann whose work in the 1940s described storing both instructions and data in the same memory. This is the foundation of the vast majority of computers made since then, from desktop PCs to smartphones.
The main components inside the CPU are: the ALU (performs arithmetic and logic operations), the control unit (directs data flow and manages the fetch-execute cycle), the clock (generates timing pulses), registers (small, fast storage locations inside the CPU) and the bus (a collection of wires for transmitting signals between components).
Inside the CPU: Key Components
The ALU carries out two categories of operation. Arithmetic operations include addition, subtraction and related calculations. Logical operations include comparisons (is A greater than B?) and Boolean operations (AND, OR, NOT). Every time an app performs a maths calculation, checks a condition in a loop, or compares two values, the ALU is doing the work.
The control unit does not process data itself. Instead, it orchestrates everything else — sending control signals to tell the ALU to start calculating, telling memory to send data, and directing results to the right place. It decodes each instruction and determines which components need to act and in what order.
The clock generates a continuous stream of pulses. Each pulse triggers the next step in a process. The faster the clock, the more operations the CPU can perform per second. Modern consumer CPUs typically operate at several gigahertz — billions of clock pulses per second.
Registers are tiny storage areas built directly into the CPU. They hold data or addresses that the CPU is currently working with. Because they are physically on the processor chip itself, registers can be accessed far faster than main memory. You do not need to know the names of individual registers for this course.
A bus is a set of parallel wires that carry signals between parts of the system. Different buses carry different types of information: data, addresses and control signals. Data travels along the bus in parallel, with each wire carrying one bit at a time.
Cache sits between the CPU and main memory (RAM). When the CPU needs data, it checks the cache first. If the data is there (a “cache hit”), it is retrieved much faster than fetching from RAM. If not (a “cache miss”), the CPU fetches from RAM and copies the data to cache for next time. Cache is much more expensive per byte than RAM, so systems have relatively small amounts of it — typically measured in megabytes. Many modern CPUs have multiple levels of cache (L1, L2, L3), each larger but slightly slower than the last.
The Fetch-Execute Cycle
The CPU works by repeating the same three-stage process continuously for every instruction in a program. This is the fetch-execute cycle:
1. Fetch — The next instruction is retrieved from main memory and brought into the CPU. The CPU knows where to look because it tracks which instruction comes next. The instruction travels to the CPU along the data bus.
2. Decode — The control unit interprets the instruction to work out what operation is needed and what data is required. Different instructions require different actions — the control unit figures out which components need to be involved and what they must do.
3. Execute — The instruction is carried out. This might mean the ALU performs a calculation, data is moved between memory locations, or the result is written back to main memory. Once execution is complete, the cycle repeats for the next instruction.
This cycle repeats billions of times per second in a modern processor. A common exam error is writing only two stages (“fetch and execute”). There are always three stages: fetch, decode and execute. The decode stage is separate and must be stated.
The fetch-execute cycle has three stages: Fetch (instruction retrieved from memory), Decode (control unit interprets what the instruction means), Execute (instruction is carried out, which may include reading or writing to main memory). This repeats for every instruction.
CPU Performance Factors
Three factors have the greatest effect on how fast a CPU can process instructions:
Clock speed — measured in hertz (Hz), typically gigahertz (GHz) in modern CPUs. A higher clock speed means more fetch-execute cycles per second. Doubling the clock speed roughly doubles the number of instructions the CPU processes per second, all else being equal.
Number of cores — a core is an independent processing unit. A dual-core CPU has two cores; a quad-core has four. Multiple cores allow the CPU to work on different tasks simultaneously (parallel processing). Modern phones and computers typically have 4 to 16 cores. More cores improve performance for programmes written to take advantage of them.
Cache size — a larger cache stores more frequently used data closer to the CPU, reducing the number of slower RAM accesses needed. Applications that repeatedly access the same data (such as games or video editing) benefit most from larger cache.