How Does a Processor Work? Intended for Secondary Schools

The processor, also referred to as the central processing unit (CPU), is the most fundamental and the most important part of a computer. It is the part that executes all tasks (instructions) that are fed to the computer through program code. At the same time, peripheral devices, RAM and storage (hard drives, SSDs, oreven tape in the past) are connected to it in various ways. It is therefore truly the central element of a computer.

Role of the CPU

The processor, referred to as the central processing unit (CPU), is sometimes simplistically compared to the brain of the computer (sometimes the comparison to a heart also appears, but this can be misleading, because one could object that in a computer, the function of blood circulation corresponds more closely to the electrical power supply, while the core function of a computer is more akin to thinking). The comparison to a brain isn’t perfect either, because the brain also handles long-term and short-term memory—whereas in a computer, these functions are handled not by the central processing unit, but by storage and RAM.

In computer science, the processor (CPU) is the component of the computer that executes the program—that is, the individual machine instructions that make up the program. At the same time, it must handle the inputs and outputs of data that control the program or with which the program works or which it produces. The CPU is therefore the most important element forming any computer and the most important part of its functionality (though not the entirety; the storage and operating memory functions, as well as the physical devices that facilitate various inputs into the computer and outputs out of it, are still needed).

First processors built from discrete components and the microprocessor revolution

Processors or electronic circuits corresponding to their function were initially made up of individual components in historical computers. At the very beginning, they were based on vacuum tubes and relays (computers built from these were very large and less reliable), later on semiconductor transistors. Even a processor composed of separate transistors or later small and medium-scale integrated circuits with multiple transistors in one package (TTL) could take up a considerable size overall and could be composed of several printed circuit boards.

What we imagine a processor to be today appeared under the name microprocessor at the beginning of the 1970s. A microprocessor is a processor that is entirely manufactured as an integrated circuit, i.e., a chip. Ideally a single one, but it can also be a few chips (interestingly, in the newest CPUs, the construction from multiple so-called chiplets has started to reappear). The first microprocessor is usually considered to be the rather limited 4-bit Intel 4004 for calculators from 1971. However, at the same time, several candidates from various companies were under development, and Intel 4004 was simply the first to market.

Microprocessors were revolutionary for the development of computers and all advanced electronics and technology derived from them. For some time in the 1970s and early 1980s, the use of processors built from discrete components persisted in more powerful, so-called mainframe computers and minicomputers, until the capabilities and performance of microprocessors (which was initially inferior) surpassed the possibilities of TTL-based processors after this transitional period.

Over time, however, continuous improvements and progress in microprocessors, driven by advancements in manufacturing technologies, enabled the rapid development of sophisticated architectures, opened up the possibility of using an ever-increasing number of transistors, and generally achieving better and better performance. All this while simultaneously reducing cost and simplifying the manufacturing of the entire computer. Microprocessors therefore gradually became the only form of processors used during the 1980s. All today’s processors (CPUs) are microprocessors, but we usually no longer use this longer term.

Processor and cores

The historically developed concept of a computer worked with a processor that processed one program (process), i.e., in a single thread—so-called serially. However, computers with multiple processors (each in its own socket on the motherboard) running simultaneously were gradually developed, allowing multiple threads (processes) to run simultaneously, i.e., in parallel. Multi-processor systems are still encountered in servers, designated for example “2S”, “4S”, “8S” according to the number of processors (the letter S stands for socket—these designations literally state the number of processor sockets on the computer’s motherboard).

In the early years of the 21st century, dual- and multi-core processors appeared. In personal computers, this was in 2005 with AMD Athlon 64 and Intel Pentium D processors. A processor with multiple cores means that within a single unit designated as a processor, several CPUs/processors are actually integrated in parallel, which can simultaneously process their own processes/threads. This is essentially a merger of the earlier multi-processor system into a single processor that occupies just one socket on the motherboard.

These individual processors have since been called cores, and the term “processor” today refers to the whole unit that aggregates all the cores. Again, it can be said that today almost all standard processors are multi-core; exceptions would only be found in more specialized deployments (embedded devices and processor cores in microcontrollers).

From microprocessor to “SoC”

In addition to the transition from earlier purely single-core processors to today’s multi-core ones, processors have undergone another transformation starting roughly at the beginning of the 21st century, evolving from a component that implemented only the functionality traditionally associated with the concept of a processor (in the sense of the CPU, that central processing unit) into a chip that integrates many other functions that were gradually added to it.

Controllers for various interfaces and connectivity which used to be provided by an external chipset began to be moved onto the CPU chip (starting with the memory controller), as did entire peripherals, such as a graphics adapter and accelerator (i.e., a GPU), or network adapters.

This creates a type of processor called an SoC (system on a chip). This abbreviation means that a single processor (chip) integrates all or almost all of the functionality that a computer requires. This definition usually does not include storage and RAM in SoCs; the processor typically does not contain these components, so in practice, to build a computer you need not only the SoC, but also RAM and an SSD, as well as some motherboard or another way to connect them.

The SoC designation is most commonly used today for mobile processors with ARM architecture in phones and mobile devices, but current x86 processors in personal computers commonly cover practically the same complete computer functionality these days and can also be considered full-fledged SoCs, although the term is used less frequently for them.

The article continues on the next page…

⠀