Secondary Storage and Embedded Systems
Why Is Secondary Storage Needed?
Main memory (RAM) is fast and directly accessible by the CPU, but it has a critical weakness: it is volatile. Switch the computer off and everything in RAM disappears. This is why computers also need secondary storage — a separate, permanent place to keep files, applications and the operating system itself.
Secondary storage provides three things that RAM cannot: persistence (data survives when the power is off), capacity (far more data can be held — measured in tebibytes (TiB)) and portability (storage devices can be moved between computers). The CPU cannot access secondary storage directly — data must first be copied into RAM before the processor can work with it.
Secondary storage is needed because RAM is volatile. It provides persistence (data retained without power), capacity (far more storage than RAM — modern drives hold multiple tebibytes) and portability (devices can be transported and plugged into other systems). The CPU must load data from secondary storage into RAM before it can process it.
Magnetic Storage
Magnetic storage encodes data using magnetic fields. In a hard disk drive (HDD), one or more metal platters spin at high speed, coated with a magnetic material. A read/write head floats just above the surface, magnetising tiny regions to record 0s and 1s. To read data, the head detects the magnetic orientation of each region as the platter spins past.
HDDs offer high capacity at relatively low cost per gibibyte — making them popular for desktop computers and large-scale data storage. However, because mechanical parts must physically move to the right position, access times are slower than solid-state alternatives. HDDs are also vulnerable to physical shock: a jolt while the drive is spinning can damage the head or platters.
Optical Storage
Optical storage uses a laser to read and write data. The surface of a disc contains millions of microscopic pits and flat areas (called lands); the laser bounces off these differently, and the detector interprets the pattern as 0s and 1s. Common optical formats are:
- CD — typically around 700 mebibytes (MiB) capacity. Used for audio and small data files.
- DVD — typically 4.7 to 8.5 gibibytes (GiB). Used for films and software distribution.
- Blu-ray — typically 25 to 50 GiB or more per disc. Used for high-definition video and large data archives.
Optical discs are robust and cheap to manufacture in bulk, making them good for distributing content. However, capacity per disc is limited compared to HDDs or SSDs, and they can be scratched or damaged.
Solid-State Storage
Solid-state storage (SSD) holds data in electronic memory chips. Because there are no spinning platters or moving heads, SSDs are considerably faster than HDDs — a file that takes several seconds to open from an HDD may open almost instantly from an SSD. They are also silent, use less power, and are far more resistant to physical shock.
The main disadvantage of SSDs is cost: they are more expensive per gibibyte of capacity than HDDs. However, prices have fallen significantly, and SSDs are now the standard in laptops, tablets and smartphones.
Magnetic (HDD): High capacity at low cost per GiB — HDDs are still cheaper per GiB than SSDs. Slower access due to mechanical movement. Vulnerable to physical shock. Produces some noise and heat. Good for large-capacity desktop storage or data centres where cost matters more than speed.
Optical (CD/DVD/Blu-ray): Low cost to produce in bulk. Portable and can be read on any compatible drive. Limited capacity per disc. Slow to write. Prone to surface scratches. Best suited for distributing content or long-term archival where the discs are not frequently written to.
Solid-state (SSD): Fast access and data transfer. No moving parts — silent, light and shock-resistant. Lower power consumption. More expensive per GiB than HDD. Best for laptops, mobile devices and applications where speed and durability matter (e.g. gaming, video editing on the go).
The key trade-off: magnetic (HDD) gives the most capacity for the lowest price but is slow and fragile. Optical is cheap to manufacture and portable but limited in capacity. Solid-state (SSD) is fastest and most durable but costs more per gibibyte. Exam questions often ask you to justify a choice for a specific scenario — match the storage type to the use case.
Embedded Systems
Most computers are general-purpose: you can install any software and use them for many different tasks. An embedded system is different. It is a computer (with a processor, memory and software) built inside a larger product to control a specific function. The user cannot install new programs on it — it does its one job, repeatedly, for the life of the product.
Embedded systems are everywhere. A washing machine contains an embedded system that controls the water temperature, drum speed and cycle timing. A smart thermostat uses an embedded processor to monitor temperature sensors and switch heating on or off. A car’s engine control unit (ECU) is an embedded system that reads sensor data from the engine dozens of times per second and adjusts fuel injection, ignition timing and other variables in real time.
Because an embedded system has a fixed purpose, it can be optimised in ways a general-purpose computer cannot. The processor can be smaller and cheaper. The software is stored in read-only or flash memory and starts running the moment power is applied — no operating system needs to “boot up” in the way a desktop PC does. The system only needs enough memory and processing power for its specific task, keeping cost and energy consumption low.
Reliability is critical for embedded systems. A washing machine controller that crashes mid-cycle is a serious problem; a car’s ECU failing at speed could be dangerous. Embedded software is therefore typically simpler and more rigorously tested than general-purpose software, with fewer features and less opportunity for unexpected states.