CPU performance, instruction sets and embedded systems

Last time

Before we start: last time, in The fetch–decode–execute cycle, you learned these points. Here they are again on one card. Keep them in mind, because today builds on them.

Why is one computer faster?

Two laptops can look the same, but one runs games smoothly and the other struggles. The difference is the CPU. Three things decide its speed: the clock speed, the number of cores, and the size of the cache.

Clock speed

The CPU has a clock that sends out regular pulses. Each pulse is one clock cycle. Clock speed is the number of cycles every second, measured in hertz. One gigahertz is a billion cycles a second. A higher clock speed means more fetch decode execute cycles every second. So more instructions are executed every second, and the computer is faster.

Cores

A core is a complete processing unit, with its own control unit, ALU and registers. A dual-core CPU has two, a quad-core has four. More cores can process more instructions at the same time, because each runs its own cycle. But doubling the cores does not always double the speed: the software must be able to share the work.

Cache size

Cache stores the data and instructions the CPU uses most often. A larger cache can store more of them. So the CPU reads from slow RAM less often, and performance improves.

Comparing processors

Here are three processors. B has the fastest clock, so it is best for one program that cannot use many cores. C has twice as many cores and the largest cache, so it is best for running many programs at once.