Performance is about the whole system

Computer performance means how quickly and smoothly a computer can complete the work it is given. It is tempting to judge performance using one headline number, such as processor speed, but real performance depends on several parts working together. A faster processor can still spend time waiting if data arrives slowly from memory. More memory can help with large programs, but it does not automatically make every calculation faster. A computer used for video editing, gaming, browsing, database searching or scientific simulation may be limited by different parts of the system.

The Higher course focuses on four key hardware factors: number of processor cores, clock speed, cache size and bus width. In exam answers, the strongest responses name the factor and explain the effect. For example, saying "more cache improves performance" is a start, but it is not enough. A better answer explains that cache stores frequently used instructions and data close to the CPU, so the processor spends less time waiting for slower main memory.

Clock speed and cores do different jobs

Clock speed is measured in hertz. A 3 GHz processor has three billion clock cycles per second. This does not mean every instruction takes exactly one cycle, but it gives a useful idea of timing: a higher clock speed can let the processor carry out more cycles in the same amount of time. Clock speed is especially relevant when a task cannot be split easily. For example, one long calculation that must be completed step by step may benefit more from a faster core than from many extra cores.

Cores are different. A CPU with several cores has multiple processing units available. This can improve performance when a computer is running several programs at once, or when one program is designed to divide work into threads. A video export, for example, may be split into chunks so different cores can work on different frames or sections. However, more cores do not guarantee faster performance for every program. If a program uses only one thread, most of the extra cores may sit idle for that task.

Cache reduces waiting time

The processor is much faster than main memory. If the CPU had to fetch every instruction and piece of data from RAM, it would often be waiting. Cache helps by storing copies of frequently used instructions and data close to the processor. When the CPU needs something, it checks cache first. If the item is present, it can be accessed much faster than going to main memory. This is called a cache hit. If it is not present, the CPU must fetch it from main memory, which is slower.

Larger cache can improve performance because it can hold more of the data and instructions that are likely to be reused. This is particularly useful in programs that repeatedly access the same data, such as loops, games, image processing and database operations. Cache is not the same as main memory: it is much smaller, much faster and more expensive per byte.

Bus width and bottlenecks

The data bus transfers data between the processor, memory and other components. Bus width is the number of bits that can be transferred at once. A 64-bit data bus can transfer more bits in one transfer than a 32-bit data bus. This can improve performance because fewer transfers may be needed to move the same amount of data. The address bus is different: its width affects how many memory locations can be addressed, not how many data bits are moved in one transfer.

Performance is often limited by a bottleneck. If the CPU is very fast but the data bus is narrow, the processor may wait for data. If the data bus is wide but the program cannot use multiple cores, adding more cores may not help. If the computer has little cache, the processor may repeatedly fetch from slower memory. Good performance explanations therefore link the hardware feature to the work being done.

Choosing the right explanation

Exam questions often describe a situation and ask why one computer might perform better than another. Read the task carefully. If the scenario is multitasking, video rendering or running many server requests, number of cores may be relevant. If it is one calculation-heavy task, clock speed may be more relevant. If the program repeatedly accesses the same instructions or data, cache may matter. If large amounts of data are being transferred, bus width may be the key point. The best answers do not list every factor; they choose the factor that fits the scenario and explain the link clearly.