【Central Processing Unit】 What is it + How does it work ▷ 2024

Central Processing Unit or CPU: What is it

The Central Processing Unit of a team is the most important component of it. It is mainly responsible and as indicated by its name, to process any type of information and to offer those ports that are completely necessary so that peripherals can be connected to it.

The CPU, as it is called in the world of computing, is made up of a series of components that together make the computer work properly. These micro units are the arithmetic-logic unit (UAL) and the control unit (UC). The union of both is what is known as microprocessor, a processor (or, rather, a set of dozens of processors) of reduced dimensions.

In this text we will try to solve all the doubts that arise about the concept of Central Processing Unit, the characteristics of its components and its operation. In addition, we are going to define the known CPUs and other curious data about this concept, such as how it differs from a graphics unit or GPU .

Index:

What is a CPU and what is it used for in computing?

As initially indicated, the Central Processing Unit (processor, CPU or CPU) is the most important part of a computer and other devices that include it, for example, a smart TV or a smartphone. In fact, it is the place where all procedures related to information are carried out .

This has a structure based on an integrated circuit called microprocessor, which can change and in fact it does between the different brands of computers that exist in the market.

This is a device that interprets a series of instructions that are contained within a program, or, that are defined in the inputs and start the execution of the data process.

It fulfills different functions:

Program execution . It ensures that their instructions are addressed for startup and commissioning.
Programming . It refers to the understanding of all the functions of the system through a previous organization chart preparation. This avoids saturation of the system and controls what it can support at all times at the disk and memory level.
Communication with all input-output units . The agreement is made between each of the units, so that peripherals can respond without a collapse between them and, usually, immediately.
Primary storage . This is equivalent to a good memory management, so that it is controlled with the programs they have the necessary space so that they can function properly, assigning it to each one that they need depending on their characteristics and the use that they are being given at all times .

Evolution of the CPU concept

In the first computers that existed, the central processing units were designed to work with a much larger computer. For this reason, the machines were operated with very simple, mechanical mechanisms. In fact, at first, these computers had to be rewired so they could perform different tasks.

The first computers with a CPU were the computers with a stored program . This was established with the ENIAC model created in 1945 and whose architecture was prepared by the mathematician who has marked the computer world, John Von Neumann.

What was intended with this equipment was to perform operations of different types that were stored in the computer memory, which meant that there were no physical changes to be made in computer wiring to avoid wasting time and effort and causing errors.

Initially, custom central processing units were created, but this is not too operational and, for this reason, it has long been considered the standardization of processors to suit various purposes . This was made possible thanks to the appearance of the integrated circuit (IC), which has allowed the CPUs to be increasingly complex in small spaces.

We must get to the idea that today microprocessors are everywhere, from cars and trucks to mobiles or children’s toys. The start of improvements for this type of units began with the appearance of the transistor in the 50s . With it you could create more complex CPUs on printed circuit boards.

Those based on integrated circuits are referred to as small-scale integration devices (SSI). These contained transistors always multiples of ten. The complete ones required thousands of individual chips, but consuming the minimum space .

In 1964, IBM introduced the System / 360 architecture , which allowed a number of computers to run the same program at different speeds. Therefore, although the computers were incompatible with each other, the architecture was the same. At that time, it was that the transistors allowed the processor to operate at much higher speeds for the switching time it offered.

Over time, the concept of microprocessor for processors manufactured with a small number of integrated circuits appeared; usually only one. The decrease in size made switching times faster. For this reason, the design, complexity and size has not changed much over the years once this microelement appeared.

Computer CPU components and features What hardware parts make up?

The Central Processing Unit consists of a series of components that we will define separately to know the scope of its functionalities:

Control Unit

This is one of the blocks that make up the CPU. Its main function is to find instructions in memory, decode and execute them through the Process Unit. In other words, UC is all the set of circuits that control the flow of data that occur in the processor.

There are many components found within this block. These are:

Instruction log . Place where the instruction is stored during the time it is being executed.
Program counter . Contains the memory address of the next instruction to be executed.
Controller and decoder . Interpret the instruction by extracting the code of the operation to be performed.
Clock . It provides electrical impulses at intervals with certain constancy.
Sequencer . It generates micro orders that are necessary for an instruction to be executed.

Process Unit

This is another of the most important blocks of this element. Its function is to execute all the tasks passed to it by the control unit .

To carry out these processes, you need the help of the elements described below:

Floating point unit . It is used to perform mathematical operations with real numbers.
Arithmetic-Logic Unit (ALU) . It is used to perform basic arithmetic operations, logical functions, comparisons or bit rotations.
Status register . Save some indicators on the result of the operations that have been done.
Accumulator record . Preserve the operands and the results of the operations.

Records

A record is a high-speed memory built into the processor itself. It is used to keep data temporarily and to be able to access the most used values . In fact, it is the best way for the system to conserve data. The registers are measured by the number of bits that they allow to store. If your device has, for example, 64 bits, this indicates that 64-bit registers are allowed to be stored.

The types of records that exist are as follows:

Floating point records . They save data in the form of a floating point.
Constant records . They have values that are created for use in read-only mode. They are made when the hardware itself is created.
Data records . They keep whole numbers.
Memory records . They only store memory addresses.
Specific purpose records . They keep information about the state of the system, such as the pointer of a stack.
General purpose records . They retain both data and addresses. They are used mainly in the architecture of Von Neumann which, on the other hand, is the typical one.

Bus in and out

The bus is used to transport data between components of a computer. They are clues about an integrated circuit that can be transferred in two ways:

Parallel . The bus allows several bits to be transmitted at the same time.
Series . The bus can only transfer the data bit by bit through a single information transmitter cable.

In the first computers, buses always worked in parallel, but in recent years they have been replaced by serial buses. Although the latter are more difficult to implement, the transfer speed they get is much higher.

How does a CPU work?

This performs multiple operations, but what stands out most of all is the program execution . The programs, seen from the inside, are a set of instructions that are represented by numbers that are stored in the computer memory before they are executed. All systems that have Von Neumann architecture execute four steps to perform their operations. These are read, decode, execute and write .

FETCH (Read)

During this step, the instruction represented by a sequence of numbers is collected from memory.

The Program counter (PC, not to be confused with personal computer or Personal Computer) is responsible for storing the memory address where these instructions are.

After reading, the PC is incremented to change the position and be located in the memory unit where the address of the next instruction is located to be executed. In this case, the slow memory is used and its response speed is related to the cache and the processor of the equipment.

DECODE (Decode)

Once we have the instruction, this is divided into different parts that will be interpreted by the units of the processor. A part of the numbers indicates the operation to be performed (opcode) and the following parts correspond to information related to the execution of said operation. For example, in the case that the instruction is a sum, it will be necessary to understand the operator that indicates sum and the operands that are intended to be added.

These can be defined by a value or by a memory address, where the data to be added is taken. In more complex CPUs, a microprogram has been created that translates the instructions , so that decoding becomes easier.

EXECUTE (Execute)

At the moment this step arrives, the processor connects with its units involved in the instruction so that the operation requested in the program can be carried out.

When the addition operation mentioned above is performed, the arithmetic-logic unit (ALU) is connected to perform the operations, with the inputs and outputs where the numbers to be added are found and the place where the result is placed.

WRITEBACK (Write)

In this step, only the writing of the results of the executed instruction occurs. These can be kept in volatile memory or kept in main memory , which can be a bit slower.

After completing the last step, the process is repeated with the next instruction cycle to read the next one and increase the program counter. In complex processors, you can read multiple instructions one at a time so you can then execute them.

How many types of Central Processing Units are there?

There are several types in terms of the Central Processing Unit, which are classified according to the number of cores it has:

CPU of a core

It is called core CPU. It appeared with the beginning of computing concepts. It was a processor with a single core, where the execution of basic tasks was very slow , performing one by one, although by that time, as you can imagine from what we have already said, it was a team revolutionary that allowed to reduce the percentage of errors as well as the derivation of resources to other needs.

Dual core processors

It was a breakthrough in the world of technology: combine two cores within a processor ; It was not an easy task. The overload in the management system must be controlled, which does not allow doubling the speed of a processor.

They were used in multitasking environments since, in it, programs and threads fight for processor time. If there is a second core, one of the threads would run on one computer and the other on the other. This UCP offers a 75% higher performance than a single core processor , not being able to talk about a double performance. Of course, it works two tasks and does it faster.

Quad-core CPU

This is a type of processor called quadcore. This is made up of four cores and is much faster than the previous two models. It is capable of performing several tasks at the same time and at great speed.

It is usually the type of processor that we find in mobile phones of medium and medium-high range.

six- and eight-core CPU

These are the ones currently used in computers. They are able to perform a lot of tasks of enormous demand at an amazing speed . Depending on their capacity, they can perform between six and eight main tasks at the same time without the user experience being dimmed.