Types of Processor

AQA 4.6.1 — Types of processor

Retrieval: the previous two lessons assumed a single, general-purpose CPU and asked how to make it faster. This lesson asks a different question: is a general-purpose CPU even the right kind of processor for every job?

Key vocabulary

• General-purpose processor — designed to handle a huge variety of different tasks reasonably well (a typical desktop/laptop CPU).
• GPU (Graphics Processing Unit) — contains hundreds or thousands of simpler cores, built for doing the same calculation on masses of data simultaneously.
• Embedded processor — a small, low-power, often single-purpose processor built into a specific device.
• Multi-core processor — a processor containing several independent processing cores on one chip.

Understand — general-purpose vs specialist is a genuine trade-off

A general-purpose CPU is built to be reasonably good at almost anything — web browsing, spreadsheets, games, background tasks — by being flexible. A specialist processor gives up that flexibility deliberately, in exchange for being dramatically better at one specific kind of workload. Neither approach is objectively superior; the right choice depends entirely on what the device actually needs to do, repeatedly, for its entire lifetime.

Visualise — matching processor type to task

Explain — why GPUs parallelise so well

A CPU core is built to handle complex, varied, often unpredictable instructions one after another very capably. A GPU instead has hundreds of much simpler cores, each of which is far less capable individually — but graphics rendering (and similarly, training an AI model) means doing the exact same simple calculation on millions of individual pixels or data points, completely independently of each other. Splitting that identical, repeatable work across hundreds of simple cores running in parallel finishes it far faster than one complex core working through it one piece at a time — the workload itself is what makes parallelism pay off here.

Apply it — classify the processor type

For each scenario, classify which processor type is most suitable and justify your choice in one sentence:

1. A smart doorbell that must run one dedicated task continuously for months on battery power.
2. A desktop PC used for everyday browsing, office work, and occasional gaming.
3. A workstation rendering a feature-length 3D animated film.
4. A server handling hundreds of independent user requests simultaneously.

(1: embedded processor — one fixed task, must minimise power draw over a long unattended lifetime. 2: general-purpose CPU — task variety is exactly what it's built for. 3: GPU — rendering is the same lighting/shading calculation repeated across millions of pixels. 4: multi-core CPU — hundreds of genuinely independent requests map naturally onto separate cores.)

Common mistake

Assuming "more cores is always better," regardless of the task. Many everyday tasks (a single web page loading, a simple mobile app) are inherently sequential and simply cannot be split across extra cores — those extra cores sit idle no matter how many are added, so they add cost and power draw without adding real-world speed for that specific task.

Challenge

Pick one embedded device you use or have seen (a microwave, a car's engine-management system, a fitness tracker). Explain, using the ideas above, why it almost certainly uses a small embedded processor rather than a full general-purpose desktop CPU, considering both cost and power draw.

Looking ahead: the next lesson (Primary Memory) moves from what processes data to what stores it while it's being used — and why a computer needs several different kinds of memory, not just one, for reasons that turn out to be strikingly similar to why it needs several kinds of processor.