The central processing unit or CPU is one of the most important components we can find in almost all modern hi-tech parts. However, most of us have rather poor notions of what they do and how they do it, how they have become the complex technological miracles they are, and what are the main modern types.
For computer engineers and racing purposes, there is probably not much more to know about how Computer Hardware and other modern gadgets work. However, for the average amateur insurance there is still something to know and I would like to finish clarifying.
For these cases in Gizmologia we have prepared a series of articles in which we will try to detail the most important aspects of the different components that give life to all those devices that help us to enjoy a higher quality of life.
What is a CPU?
Although it can not be said that there is a single most important part inside a computer, since more than one is totally vital for its operation, the CPU or processor can be considered the cornerstone of these machines. And it is precisely this component that is responsible for computing, ordering or processing, concepts that define modern computers or computers.
At the moment they are complex pieces of technology developed with microscopic architectures, most of them comes in the form of a single chip, quite small, from there they were called microprocessors some decades ago.
Today the processors are in virtually every object we use in our day to day, such as televisions, smartphones, microwave ovens, refrigerators, cars, sound equipment, and of course, personal computers. However, it was not always the small miracles of the technology they are now.
A little history
There was a time when the processors were made up of huge armatostes that could well fill a room. These first steps of computer engineering were mainly constituted by empty tubes, which although at the time were substantially more powerful to the alternatives formed by electromechanical relays, today the 4 MHz that at most they used to reach would seem to us of laughter.
With the arrival of the transistors in the 50s and 60s, the CPUs began to be made, in addition to being smaller and more powerful, also much more reliable, since the machines formed by vacuum tubes tended to present an average failure every 8 hours.
However, when we talk about getting smaller we do not mean that they fit in the palm of the hand. Just went from being room to fridge size. And those still large processors consisted of dozens of printed circuit boards that were connected to each other to give life to a single CPU.
After that came the invention of the integrated circuit, which basically linked everything in a single printed circuit board or wafer, which was the first step to the achievement of the modern microprocessor. The first integrated circuits were very basic, since they could only group a handful of transistors, but over the years, and the exponential growth of the number of transistors that could be added in an integrated circuit, by the middle of the Sixties We already had the first complex processors constituted by a single wafer.
The first microprocessor as such would be introduced to the market as early as 1971, it was Intel 4004, and since then the rest is history. With the rapid evolution of these small chips, and their great flexibility, they have completely monopolized the computer market, being that except for very specific applications that require highly specialized hardware, they are the nucleus of almost all types of modern computer.
How a CPU Works
Simplifying to the extreme, and in didactic terms, the operation of a processor is given by four phases. These phases are not necessarily always separate, but usually overlap, and always occur simultaneously but not necessarily for a specific function.
During the first phase the processor is responsible for loading the code from memory. In other words read the data that must be processed later. In this first phase there is a common problem in the architecture of the processors, and is that there is a maximum of data that can be read by period of time, and are usually inferior to those that can be processed, reason why there is a species Of effect of peak of bottle that at the moment is tried to solve applying the multichannel and cachés.
In the second phase the first stage of processing as such occurs. The information read in the first phase is analyzed following a set of instructions (next section of this article). Thus, within the data read there will be descriptive fractions for the set of instructions, which will indicate what to do with the rest of the information. To put a practical example, there is code that indicates that the data of a package must be added with those of another package, each package being information that describes a number, whereby a common arithmetic operation is obtained.
Then comes the phase that continues with the free processing, and is responsible for executing the instructions taken decoded within the second phase. In the example above, this would be where the sum is done and the result is obtained.
Finally, the process concludes with a writing phase, where again the information is loaded, only this time from the processor to the memory. In some cases the information may be loaded into the processor memory to be reused later, but once the processing of particular work is completed, the data always ends up being written to the main memory, where it can then be written to the storage unit Or not, depending on the application.
Major modern architectures
As we have already said, the function of the processor is to interpret information. Data is loaded from different memory systems in the manner of binary code, and it is that code that must be processor to be converted into useful data by the applications. The said interpretation is realized through a set of instructions, which is what defines the architecture of the processor.
Currently, two RISC and CISC architectures are mainly used. RISC gives life to processors designed by the British firm ARM, which with the rise of mobile devices has seen significant growth. Also PowerPC, the architecture that gave life to Apple computers, servers and the Xbox 360 and PlayStation 3 consoles, is based on RISC. CISC is the architecture used in AMD’s Intel and X86-64 X86 processors.
Regarding which architecture is better, it has always been said that being much cleaner and better optimized RISC would be the future of computing. However, Intel and AMD have never given the arm to twist and have managed to create a very solid ecosystem around their processors, which although being heavily contaminated with obsolete backward compatibility elements, have always kept up with their rivals.
In general, due to its flexibility and relative ease of production, for some years more CPUs will remain the centerpieces of modern computing. But we must always keep in mind that with the passing of the years have been evolving technologies in parallel that help to decentralize the load and today more than ever graphic processors, more powerful but less flexible have begun to gain almost as much importance as the unit Of central processing.