What is processor in computer? Over 95% of you viewing this article today must have at least one, and even more than a single computer (laptops, desktops, etc) of some kind in your homes.
A majority of you may have a lot of knowledge about these devices, and it is okay if others don’t. But all of you, at some point must have thought, what makes a computer capable of computing?
What is that one component without which you would not be able to call the set of all the other components like monitor, RAM, etc, together as a computer?
Thereby, we present to you the heart and soul of any computing device, the processor. Check out Intel Atom, Celeron and Pentium Processors.
What is a Processor?
A processor is a silicon-based chip that has a complex internal electronic circuit and performs all the work that you do, no matter how simple or how complex on a computer.
A processor or a CPU (Central Processing Unit) situated in the motherboard of computers has another term and is called the microprocessor.
This is the fundamental workforce of a PC, no matter what size or kind, and all those other devices that are capable of computing.
It controls all sorts of tasks that a PC does, including data transfer and processing, running applications and software, input/output, and the working of all other peripheral components that you use and so on.
But it doesn’t work alone, and there are other sensitive parts that you need for a processor to work, like RAM, storage, OS, etc.
The CPU has multiple parts like frequency, the number of cores, and more others like them that you must know of.
Not only for general knowledge but also because a little idea about those will make it easier to buy or choose a processor later.
We will talk about each of these features in detail in another article later. For now, the main parts of a CPU are:
- Arithmetic Logic Unit or ALU: This part of the CPU is responsible for all the logical calculations, or is simply the processing sector. While it is a very small complex part, the work it does for a CPU is massive. It has several logical gates, multiplexers, etc, and takes input as binary to provide the necessary output.
- Control Unit: This is the other principal component that controls how the computer is to respond to the instructions that are sent to the processor. It works as a bridge between the other two components, namely the ALU and the memory unit, and transfers data between these two. So the input and output department of data is handled by it.
- Memory Unit: The memory unit in the CPU necessarily stores data and information but in a limited amount. This is what you refer to as the RAM or primary memory in colloquial language that stores data for the processor to work with. It may not have any role in processing directly, but without the CPU would never be able to work either. But faster RAM in a computer does give it a performance boost.
Every CPU therefore mainly does two types of tasks, among countless others:
- Arithmetic operations: Addition, division, subtraction, multiplication, AND, OR, NOT (and the other related logical operations).
- Data Transfer: Moving data from one part of the CPU to a specified memory area, and also the opposite.
Types of Processor
The modern brands that make processors like Intel and AMD have different classes of processors like Intel Core and Ryzen series respectively with the different number of cores, clock speeds, etc. Here, we will be talking about the primary types of the processor itself.
So basically a processor can be of several types. This division is made by the type of architecture they are built upon, the number of working cores they have, and so on, but one of the main points of difference is how they are used.
You see, a processor can be employed in other places than only in the CPU of your PC. Know choosing the right processor for your PC.
Thus, a modern processor can be categorized as:
- Micro Controller
This is the less sophisticated type between the three and has all the features like a processing unit consisting of one or more CPU cores, storage unit, data transfer paths, and such that allows it to work inside an embedded system, or controls a part of it.
They are most useful where the tasks are limited or are pre-defined. Thus, a microcontroller is what you would expect to be on your TV remote but not on your PC’s motherboard, but it is a form of processor nonetheless.
- Micro Processor
This is created on a microchip and is closest to the definition of a general processor that is used in a computer.
They look the same as a microcontroller but have multiple advantages over them.
It is a system of integrated circuits that encapsulates the working of a CPU with numerous other functions that allow you to do various tasks.
- Systems on Chip (SOC)
This is the technology that is used in some modern PCs, laptops, smartphones, etc.
It is based on a microprocessor and a microcontroller together while having everything necessary to run an entire system on a single chip, hence the naming.
The SoC includes the CPU itself along with all the other interfaces like graphics, audio, etc.
SoC-based systems are increasingly becoming the primary format of portable computing these days since they are smaller and you would be amazed at how all of that functionality can be put inside a single chip.
Now microprocessors can be of further sub-types depending on structure and applications.
Based on application:
- General Purpose Processor
These include the microprocessor, microcontroller, and almost all other types of a processor that comes to mind when supplying a PC with an instruction set is concerned in an embedded system.
They take data in the digital form and convey the relevant output through the various data, address, and control buses.
A vital part of modern technology is media, in its various forms and standards. But every form of media needs signaling and processing of those signals to turn them into outputs that can be displayed on various screens.
Here comes the use of the Digital Signalling Processors or DSPs. They are the specialists in signal processing and work far better than a general processor at fetching and computing analog signals steadily.
Thus, they can be found in most forms of telecommunication sectors, radar, sonar, and appliances like our phones. The inherent power efficiency and lower cost have proved them to be of much utility.
- Application-Specific System Processor (ASSP)
The Application System Integrated Circuit or ASICs as they are known now serve a single, exclusive purpose only.
They are a part of every embedded system and consist of an internal circuitry to perform the relevant tasks.
For example, one that serves to light up the headlamps of a toy car. They have further divisions and are made based on different designs, but we won’t be going in such details as of now.
But know that they are closely related to SoCs. These are very useful in large scale productions.
So, the processors that use this technology of the ASICs are known as ASSPs. They work in the same way and are mostly used in the signal decoding and encoding industries.
- Application-Specific Instruction Set Processor (ASIP)
These are the ones having instruction sets designed to suit a specific application only.
Thus, rather than supporting a whole bunch of applications, they accelerate only some of the specific ones, thus are rightly the combination of flexibility of GPPs and specialization of the ASICs.
Though their uses are limited, we as consumers are greatly benefitted by such ASIPs in the areas of signaling and video/audio encoding.
Also, with lower power consumption and production costs, designers do not have to worry about making GPPs that have to cater to every specialized function.
Based on internal architecture:
Reduced Instruction Set Computing is a form of processor Architecture which is characterized by lower power consumption and simpler instruction sets.
In the RISC models, the microprocessor works with a simple set of instructions only, and the focus is given on the software of the device in use.
Each set of instructions is completed within one machine cycle, and it is easier to design processors and code based on RISC.
But it requires more lines of coding. The task of adding two digits on your computer, for example, would require several lines of code, that may be simple to create, but are long.
CISC or Complex Instruction Set Computing is, however, a different but related processor architecture that uses complex codes, and the CPU uses a single code to perform multiple low-level tasks.
These instructions may take up to ten machine cycles to be completed, and hence processors based on this design take a longer time to fulfill tasks.
The focus is given more on hardware here. The CISC approach uses lesser lines of coding but is difficult to design.
Both the CISC and RISC models had their advantages and problems, but neither is solely used in computers anymore.
With passing time, more technology has been incorporated that fuses these processor models, both with each other and other designs that are further consumer-friendly.
- x86 based Processors
Processors, based on the x86 architecture have been around for quite some time now since their use in the Intel 8086 processors, and are still used in modern-day consumer PC’s and servers alike.
However, they have gone through many changes, like the bus width was increased to 32 bit in 1985, and again in the early 2000’s to 64 bit. Both Intel and AMD have been using this, but Intel has the parental rights here.
Some examples would be the Atom, Pentium, Celeron and Core i3-i9 series from Intel and all the Ryzen CPUs from AMD.
- ARM-based Processors
ARM is the acronym for Advanced RISC Machines, introduced in 1985 by a private company named ARM Holdings.
It was a design developed to be adopted by various other companies to make processors for embedded systems based on this model.
The modern ARM processors released in 2011 like the Armv8-A and AArch64 have both 32 and 64-bit instruction sets and are used now because of their power efficiency and lesser heat generation tendencies than x86 processors.
They have found uses as modern SoCs in smartphones, tablets, laptops, etc from popular brands like Samsung and Apple.
Based on the number of physical cores:
- Single-Core Processor
These are the primitive processors introduced after the very invention of a microprocessor.
At that time it was wonderful how a room full of computing equipment could be replaced by such a smaller device, but with further development, it started to feel insufficient.
With a single-core, they can run only single-threaded applications at once, so forget about multitasking.
These have been widely replaced by multi-core processors at the onset of the 21st century. Some examples would be 807UE, E1, etc.
- Multi-Core Processor
Multi-core processors started popping up with IBM’s Power4. Some of the models had a max speed of 1.9 GHz, and two 64-bit microprocessors.
More cores meant more threads and thus computers were then able to perform multiple tasks at once.
But now, the scene is entirely different as AMD has released the Threadripper 3970X with 32 cores and 64 threads!
In daily use, only powerful workstations can utilize such power.
On a consumer level, one may have anything between 2 and 12 cores in their PC at home or the offices, depending on how much they can spend.
Multiple cores mean better processing, and hence more power for both single and multithreaded tasks.
Both Intel and AMD have quality CPUs in various price segments.
The lifespan of a Processor
In case you are wondering, a CPU never faces physical damage or disintegration at a cellular level.
However, several other factors can and will damage the intricate circuitry or transistors present in it if proper precautions aren’t taken.
The first among these is heat that every processor produces while operating. In the absence of a sufficient cooling system, the excess heat is harmful to not only the processor but also other components.
Then, this problem is faced more by those who overclock the CPU, thus making it work at greater temperatures than usual.
For an average user, the CPU working under ordinary conditions of temperature and voltage should last about 15 years before any major internal damage or decline in performance is noticed.
However, with the given fast trend of updating computer technology, you will need to replace a CPU long before it dies out because at some point it will feel outdated, and even obsolete.
For those who overclock, you can expect a CPU to work for about 5-10 years, depending on how you use it or to what extent you put it under overload.
How Does it Work?
Now that we have told you a bit about processors, let us now see how one works.
Well, each type of processor discussed here doesn’t exactly work in the same way, but to keep things relevant to our topic of discussion, the following is how a typical CPU works.
We are keeping things as simple as possible, and remember that there are many more complex processes that go on inside a modern CPU. Each instruction is optimized in a way so that accessing time is reduced, and as a result, the CPU works faster.
The main parts of a processor have already been mentioned. But, there are yet others that have to be specified so that its working can be understood. This includes read and write lines, transistors, data and address buses, registers, cache, clock speed, etc.
At first, you need to realize that a computer doesn’t understand any language in the form as we do. So while it will easily solve the tough calculation you have been stuck with since yesterday in a matter of seconds, it needs to be made understood about it in the machine code, or signals.
Here signals can be conveyed in the form of binary code, 0s, and 1s, where 0 indicates a low signal, and 1 indicates a high signal.
Now related to this are concepts like logic gates, multiplexers, half and full adders, etc that your math or computer classes should be able to help you with, in case you have no idea what these are.
But for now, let us say that these can be used to do various calculations like addition, subtraction, comparing two values, and so on.
So, before the calculations are done in the logic gates inside the CPU, the raw instruction has to be stored somewhere before the CPU can start processing it, and then sent to the relevant components for them to work.
This is where the memory and buses do respectively. But here memory refers to smaller, faster storages called caches and registers.
While both work with the same purpose, registers are yet smaller which immediately holds the data being processed and the location of such data in the RAM, among other instructions.
The processing of any data is done in three elementary steps known as the fetch-execute cycle.
The steps are: fetch, decode, and execute. The RAM has several addresses, each storing a particular instruction.
To execute a program, the CPU asks the RAM for a specific memory location inside it using the unidirectional address bus.
This can be done serially, or in any random order. The instructions are then loaded from the RAM into the CPU using the data bus, and these steps go on inside the CPU on a loop.
The CPU takes an instruction, decodes it, and after each execution, data is either kept inside the CPU for further processing, stored back into the RAM or storage or shown on the monitor, depending on what sort of instruction was sent initially.
The decoding of an instruction is done by converting it into signals so that the CPU can interpret it.
Every instruction has two parts, the instruction itself and the memory location inside the RAM to where it belonged.
Modern CPUs can perform billions of these operations in one second, which you know as the clock speed of the CPU. Thus, a 3 GHz processor does 3 billion operations in a single second.
Now inside the CPU, there are certain preloaded instructions known as the instruction set, the basics of which include Store, Load, Jump, Compare, In, Out, etc.
Each has its separate function, and after the execution of an instruction, the output is stored temporarily in the registers.
The Jump instruction transfers the processing from one RAM location to the other, and in case it is absent the other instructions run in a loop, being executed one after the other. All processing takes place inside the ALU.
The Control unit then directs the CPU whether and when it should be sending the data back into the RAM, and from there it is sent to the necessary components.
The processor is one of the most complex parts of any computer. We have tried to show everything necessary while also keeping things simple.
Having said so, you must be wondering how to utilize all this without knowing what to look for when buying a CPU for your PC. Check out the articles related that solve this problem of yours.