GPU

What is a Graphics Processing Unit (GPU)?

    

A graphics processing unit, also known as a graphics processor or GPU, is a type of electronic circuitry designed to accelerate computer graphics and image processing on a variety of devices, including graphics cards, motherboards, cell phones, and personal computers (PCs).

    

GPUs can perform mathematical calculations quickly, reducing the time it takes for a computer to run multiple programs, a feature that makes GPUs an important enabler for emerging and future technologies such as machine learning (ML), artificial intelligence (AI), and blockchain.

    

Prior to the invention of GPUs in the 1990s, graphics controllers in PCs and video game controllers relied on the computer's central processing unit (CPU) to perform their tasks. Since the early 1950s, the CPU has been the most important processor in a computer, executing all the instructions needed to run a program, such as logic, control, and input/output (I/O). However, with the advent of personal gaming and computer-aided design (CAD) in the 1990s, the industry needed a faster, more efficient way to combine pixels in a short period of time.

    

In 2007, Nvidia built CUDA (Compute Unified Device Architecture), software that gave developers direct access to the parallel computing power of GPUs, allowing GPU technology to be used in a much wider range of applications than ever before. the 2010s saw leaps and bounds in GPU technology, perhaps the most important of which were ray tracing (which generates a computer image by tracking the direction of light from the camera) and tensor cores (designed to enable deep learning).

   

Because of these advances, GPUs have played an important role in AI acceleration and deep learning processors, and have helped accelerate the development of AI and ML applications. Today, in addition to powering gaming consoles and editing software, GPUs power cutting-edge computing capabilities that are critical to many organizations.

     

How Does a GPU Work?

  

Today's GPUs utilize many multiprocessors to handle all the different parts of the task they are meant to perform.

    

GPUs have their own Rapid Access Memory (RAM) - a specific type of electronic memory used to store code and data that the chip can access and modify as needed. Advanced GPUs often have RAM specifically built to hold the large amounts of data needed for compute-intensive tasks such as graphics editing, gaming, or AI/ML use cases.

    

Two popular GPU memories are Double Data Rate Synchronous Dynamic Random Access Memory for Sixth Edition Graphics (GDDR6) and the next-generation GDDR6X. GDDR6X consumes 15 percent less power per transferred bit than GDDR6, but because GDDR6X is faster, it consumes more power overall. iGPUs can be integrated into a computer's CPU, or plugged into a slot next to the CPU and connected via PCI Express ports. The iGPU can be integrated into the computer's CPU or plugged into a slot next to the CPU and connected via a PCI Express port.

   

What are the Different Types of GPUs? 

      

There are three types of GPUs: discrete GPUs, Integrated GPUs, and virtual GPUs.

   

Discrete GPU: A discrete GPU, or dGPU, is a graphics processor that is independent of the device's CPU, which is responsible for receiving and processing information so that the computer can function properly. Standalone GPUs are typically used for advanced applications with specialized requirements, such as video editing, content creation, or high-end gaming. They are different chips with connectors for separate boards, which are usually connected to the CPU using a shortcut slot. one of the most widely used discrete GPUs is the Intel Arc brand, which was created specifically for the PC gaming industry.

    

Integrated GPU: An integrated GPU or iGPU is built into the infrastructure of a computer or device, usually next to the CPU. The integrated GPU design was introduced by Intel in the 2010s. Subsequently, manufacturers such as MSI, ASUS, and Nvidia realized the powerful benefits of combining the GPU with the CPU (without requiring the user to add the GPU themselves via a PCI Express slot), and so integrated GPUs became increasingly popular. Today, integrated GPUs remain a popular choice for laptop users, gamers, and others running compute-intensive programs on PCs.

   

Virtual GPU: A virtual GPU has the same functionality as a discrete or integrated GPU, but without the hardware. A virtual GPU is simply a software version of a GPU built for a cloud instance and can be used to run the same workloads as a physical GPU. In addition, because there is no hardware, virtual GPUs are simpler and less expensive to maintain than physical products.

What is the difference between a GPU and a CPU?

  

       

CPUs and GPUs are similar in design, such as having many cores and transistors for processing tasks, but CPUs are more versatile than GPUs in terms of functionality. GPUs tend to focus on a single, specific computational task, such as graphics processing or machine learning.

    

The CPU is the heart and brain of a computer system or device. the CPU receives general instructions or requests from a program or software application about a task. In contrast, the GPU performs more specific tasks, often involving the rapid processing of high-resolution images and video. To accomplish its tasks, the GPU constantly performs complex mathematical calculations required for graphics rendering or other computationally intensive functions.

    

One of the biggest differences between CPUs and GPUs is that CPUs tend to use fewer cores and perform tasks in linear order, while GPUs have hundreds or even thousands of cores and support parallel processing, resulting in lightning-fast processing speeds.

    

The first GPUs were built to accelerate the rendering of 3D graphics, making movies and video game scenes look more realistic and engaging. The first GPU chip, Nvidia's GeForce, was released in 1999 and quickly followed by a period of rapid growth that allowed GPU capabilities to expand into other areas due to its high-speed parallel processing capabilities.

      

Parallel processing or parallel computing is a type of computing that relies on two or more processors to accomplish different subsets of an overall computing task. Before the advent of GPUs, older generation computers could only run one program at a time, often taking hours to complete a task. the parallel processing capabilities of GPUs allow for many calculations or tasks to be performed at the same time, making them faster and more efficient than the CPUs found in older computers.