GPUs are Faster, Why Do We Still Use CPUs?


The power of computing centers of our computers and smartphones is increasing day by day. The two basic hardware circuits known as the Graphics Processing Unit (GPU) and Central Processing Unit (CPU) are the backbone of our fast computers. In recent years, the performance of GPUs has outpaced CPUs. So why can’t we just give up the CPUs?

The manufacturers, who made great strides in the mobile processor chip, did what was once impossible with the 7-nanometer architecture. Even the best chips that came out four years ago were twice as big as today’s chips. If we consider both the CPU and the GPU in the mobile chips, we can see that the point is far ahead of the giant processor circuits on desktop PCs. Nvidia, Intel and AMD also do not go into the work they do not extend the word. All of them develop extremely high performance external processing units.

Today, CPU, GPU, what is the difference between them, most importantly, if the GPUs are more powerful, why not use the CPU answer questions. So let’s start.

What is GPU?
GPUs are designed to be experts in one field only: Calculating and processing graph data. The majority of older GPUs used to calculate “floating points” and “moving 3D objects”. Nowadays, they can process dozens of different visual layers such as texture processing, shadow and dynamic light, antialiasing and filtering.

3D computer graphics, in particular, require exceptional mathematical computation: we live in a three-dimensional space, physically using two-dimensional displays. GPUs are rebuilding non-existent 3D spaces within these screens. It is also the job of GPUs to project the entire visual world and adapt it to the screen we use. In short, GPUs are high-processing units because they are designed to handle really hard work.

What is a CPU?
CPUs are not very good for the processes mentioned above. They are more general purpose and are designed to calculate other tasks on almost all electronic devices, especially computers and smartphones. They control the information traffic between parts, such as the coordination of parts in the system, memory devices, cache systems. Systems cannot be started without CPUs, for example BIOS boot codes on computers, and specific systems on other devices.

The CPU manages all the resources of the system. The bits on the buses handle the operations during the reaction time until a letter entered on your keyboard is reflected on your screen. So the brain is the CPU. GPU is the visual perception, processing and mirroring unit of this brain. GPUs have higher processing power and capacity because visual processing involves higher data input and more difficult problems.

So why not use GPUs instead of CPUs?
Ideas for designing and producing CPUs with GPUs’ processing power have been around for a long time. First of all it is necessary to know that such a thing will be impossible and all the architectures of computers will have to be redesigned. Starting from designing a motherboard without a base unit, we may have to write a new operating system to run this platform.

If you want to understand the fundamental difference between GPU and CPU, the following analogy can help:

GPUs are like Usain Bolt. They perform a large number of transactions in a short time. Their job is to run 100 meters in the shortest time and to give the best graphic processing results in the shortest time. CPUs are like decathlon athletes. Decathlon is a branch of 10 different types of athletics, athletes compete in all of these branches. In other words, they have to endure for long periods of time, to perform long-term high performance in different areas, and in some cases to use the resources in their bodies sparingly.

The type of process CPUs focus on and the tasks they run vary according to user demands. If we want to use CPU instead of GPU, we ask a decathlon or marathon runner to set a world record at 100 meters. For this, it is necessary to go down to that athlete’s childhood and train him as a short distance runner. In short, if GPUs are used instead of CPUs, it is necessary to redesign all computers, smartphones, from the smallest to the largest, the weakest to the most powerful, which are common today. If the new systems reach the consumer, if they succeed, it can take almost a century like computers.


Please enter your comment!
Please enter your name here