Factors Affecting Overall System Performance

The capstone hardware-performance subtopic — you now need to pull together everything from Grades 10 and 11 into concrete recommendations for improving a real system's speed and reliability.

Modular design — how a system's physical layout affects performance and flexibility:

Ports and cards vs onboard components — a dedicated expansion card (e.g. a separate graphics card) generally outperforms an onboard/integrated equivalent because it has its own dedicated processor and memory, rather than sharing system resources.
Buses — as covered in 11.1.3, the speed and width of the buses connecting components sets a hard ceiling on how fast data can move between them, regardless of how fast the individual components are.

Co-processors — dedicated secondary processors that handle specific tasks, freeing up the main CPU:

Graphics Processing Unit (GPU) — handles rendering images/video, especially valuable for 3D graphics, gaming and video editing.
Maths/floating-point co-processor — handles complex numerical calculations, freeing the CPU for general instructions.

Techniques to improve processor speed:

Technique	How it helps
Hyper-threading & multi-processing	Lets the CPU handle more instruction streams concurrently (see 11.1.3)
Increasing L1/L2/L3 cache size	Reduces how often the CPU must wait on slower RAM (reduces latency)
Increasing register size	A larger register (e.g. 64-bit vs 32-bit) can process more data per instruction and address more memory; this also affects the width of the data and address buses it's paired with
Clock multiplication / overclocking	Runs the CPU (or a component) faster than its rated speed, at the cost of more heat and reduced long-term reliability

Other ways to improve overall performance:

Increasing the speed and/or size of RAM — reduces reliance on slow virtual memory/paging (see 11.1.5) and lets more programs run smoothly at once.
Matching upgrades to the actual task — e.g. a faster/dedicated video card specifically benefits 3D rendering; a faster SSD specifically benefits video editing workflows that constantly read/write large files.

Example

A video editor complains their system is slow when exporting 4K footage. Since export is disk-and-CPU intensive, the best recommendations are: upgrade to a faster SSD (reduce read/write bottleneck), add more RAM (reduce paging), and consider a dedicated GPU with hardware video encoding support — rather than just 'buy a faster CPU', which may not be the actual bottleneck.

💡 Exam Tip

Grade 12 questions expect you to diagnose the bottleneck first, then recommend the specific upgrade that addresses it — a generic 'upgrade everything' answer scores far fewer marks than a targeted, justified recommendation.