The Basics you should know : Graphics Card (GPU)

Get the basic, detailed knowledge and working of Graphics Cards. Know why and where the Graphics Cards are used? and Whether you need it or not?

Before writing about GPU, I would like to clear one thing that GPU is the main core chip/processing unit that is embedded on the card body called as a Graphics card. Graphics card is also known as Video card. The hardware part that you are looking at in the above image is Graphics card which has GPU embedded on it.

I have seen that many people who recognize them as the card or hardware which is used for gaming only. Now the fact is- it is partly correct if we consider their use today, but this answer would have been 100% correct 15 years ago. Initially, Video cards were invented for supporting better graphics and display quality for gaming/visual performance only. The main focus was to create a good hardware device that can handle and support high video processing and rendering for games. In the early computers, framebuffers (a portion of RAM or other primary memory) had very limited memory to process continuously changing graphics/visual (as we play games we see different movements of the objects on display). Hence, after World War II (after the 1950s) as the use of computers came in demand in different fields like science and military: scientists, inventors, mathematicians, etc. started working on new technologies and devices for better quality displays and video processing. As a result, the first Video Card came into existence. Later, various companies designed and launched better Video/Graphics as the technologies and science kept improving with new findings.

Now, you know that GPU is used for gaming as it helps the computer to process all the good quality of images/frames/movements that we see on screen. The question is —

What kind of processing GPU exactly does?

I have explained about CPU in my previous story about the Processors about what it does and how it is responsible for processing everything that the computer does. GPUs have a similar task as CPUs — Process data. But GPUs are more efficient, faster and have more cores which result in accelerated performance and processing of complex data. Modern GPUs are more efficient in processing complex computer graphics. They have better parallel structure(the structure that helps in more efficient parallel processing) which makes it more efficient than most of the general CPUs for processing large block of data in parallel. As I mentioned earlier in this paragraph that GPUs have more cores(Hundreds/thousands) than the general CPUs (2 to 8). But we cannot compare them by numbers as GPUs have the more optimized architecture to do a limited set of things very well (Like image/visual data processing- handling visual data and it’s complex calculations). A CPU has more General architecture for doing a broader set of things(Every type of instruction that the computer needs to execute). The advantage of GPU is that it divides one large/complex task into smaller tasks (assigned to cores available in large numbers)and executes/performs them parallelly at the same time which makes it more efficient. Computer graphics is very complex, it needs to carry heavy and continuous calculations of data which controls/holds the information of each pixel on the screen and needs to process it efficiently and provide a smooth experience to the user on-screen.

(Optional) → Detailed Explanation for Graphics/ Visual Data Processing- The first thing you must know is that everything computer does and what we see on the screen/display is a result of code (computer program). In computers, there are special types of memories(buffers or say VRAMs →Video Random Access Memory) to store data related to display/video. These memories hold the data that is used to display the content on the screen. This data holds the information about each pixel to be displayed on the screen of the given/current resolution. Now everything that’s happening on screen is the output of something that the computer has processed based on the input. Here, the computer not only has to process the input to generate the output but also generate the representational output data to be displayed on the screen. For this, the computer has to process graphical /visual data again. Now imagine how frequently the content on the screen changes whenever we do anything like from moving a cursor to playing a video game. In games and videos- display content changes more frequently every second. As we know video is nothing but a set of images(frames) that are being displayed by changing specific number images or frames (in sequence) per second. More are the frames per second (FPS) more smooth the video is. In the case of video, the visual data is already available which needs to be just displayed after a few computations required. But in the case of gaming, it requires continuous and more complex calculations per second to process the graphical/visual data as the visual data depends on the user input (Every move we make in the game). Which makes it processing of at least 20FPS(bad performance →not so smooth) to onwards (40FPS to 60FPS means good) for acceptable performance. How many frames will be created per second is dependent on the capacity and performance of GPU that how many frames it can calculate/process each second. Can you imagine how complex this calculation is to generate the data for each pixel (smallest element of the display) of each frame from the set of frames in each second?. For such complex calculations and releasing the workload from CPUs, GPUs were created.

That’s why to handle heavy games and software we need a good Graphics card to process those high-quality graphics, textures, processing large operations parallelly, and other visual details to achieve better FPS (Frames per Second) and smooth experience.

About the term GPU from Wikipedia —

The term “GPU” was coined by Sony in reference to the PlayStation console’s Toshiba-designed Sony GPU in 1994. The term was popularized by Nvidia in 1999, who marketed the GeForce 256 as “the world’s first GPU”. It was presented as a “single-chip processor with integrated transform, lighting, triangle setup/clipping, and rendering engines”.

Modern GPUs and GPGPUs

General Purpose GPUs can perform non-specialized(not only specific tasks) calculations that means it removes the limitation(restriction)/drawbacks previous GPUs had. That means, these modern GPUs (GPGPUs) can perform and handle all the -calculations/tasks/execution/operations that are traditionally handled by CPUs. After 2007, GPGPU (General Purpose Graphics Processing Unit) highlighted and took more attention to utilize the processing powers of GPUs in handling other applications too. Modern GPUs are more efficient and can perform a large number of tasks. Today, we can run the regular application using GPUs i.e. we assign the job of handling computations for any application. As you can see, by default any general application will run using integrated (within processors) graphics processor(Intel). But I can choose to run this application with my dedicated Graphics card (Nvidia GTX 1050). The combination of CPU-GPU is very powerful if you have the right pair. Today, GPUs are used in almost every sector where high performance and efficient parallel processing is required. The fields like Science, Virtual Reality, Data Mining, Big Data, AI, Data Science, Image processing applications, Visual effects (Entertainment field for better graphics in movies), Video Editing, Graphics Design, Military, etc. and many other fields are using Graphics card. GPGPU offers us to utilize its resources for other applications.

Fun fact : For mining the Cryptocurrency, many people use a high-quality Graphics card which helps them to earn more points. Also, many hackers use GPUs to crack the passwords by running thousands/millions of matching passwords per second against the targeted database(They have set of most commonly used passwords plus the set of sample passwords generated using your personal details. They encrypt these set of passwords using the same encryption method used to encrypt your password by website or application and then they use it to find the match which gives them your password if the match is found).

Whether I need Graphics Card or not ?

As I mentioned above in this story, GPU is needed only when you or the software you want to work with needs the complex and heavy calculations. Like Gaming, Mining, processing Big Data, using Video Editing Softwares, etc. such applications need Graphics card. Not to forget that every computer comes with integrated graphics, and their quality and power depends on the CPU. Expensive and good quality CPUs comes with good quality of Integrated graphics which are generally enough to work with some level (enough for home and office usage if you are using only basic/general applications) of applications. Many Laptops and Desktops come with a dedicated graphics card where you will find many options based on your budget as well as performance.

Note: If you are buying a laptop then it is better to have at least a Graphics card that provides adequate performance because unlike Desktops, Graphics card cannot be inserted in Laptops. Graphics card are attached to the motherboard and are neither removable nor you can insert if you don’t already have it on board. But in desktops, generally, slots are provided for Graphics card and/or other hardware devices.

Note: There are some Laptops(e.g. MacBook Pro) that support eGPU(External GPU) which are connected through an external port. But these laptops and supported eGPUs are very expensive.

There are many more things about GPUs and it’s Architecture that provides exceptionally advanced features that are not needed to know at a basic level. You can always explore it on the Internet. For example, recently NVIDIA launched RTX(Ray Tracing Platform) technology which is an interface for real-time ray tracing for rendering graphics. RTX facilitates a new development in computer graphics of generating interactive images that react to lighting, shadows, reflections.

“If the CPU is King, then the GPU is his Army”

I hope you have enjoyed this article. If you have any suggestion or point, do comment below or get in touch. Thank You.

|Gmail- baderaharsh@gmail.com ||Instagram username- baderaharsh|