Shulou(Shulou.com)06/03 Report--

This article will explain in detail what the GPU parameters are, and the content of the article is of high quality, so the editor will share it for you as a reference. I hope you will have a certain understanding of the relevant knowledge after reading this article.

GPU parameter interpretation

At present, masterpieces including the movie "Avatar" and the game "StarCraft 2" continue to stimulate our eyeballs and nerves. "Avatar" allows us to see a grand scene created by 3D rendering. "StarCraft 2" makes us feel its high requirements for graphics cards under the improvement of game picture quality. If the gorgeous picture quality of Avatar and StarCraft II makes netizens gradually feel the importance of graphics cards, then with the release of windows7 system and its official support for GPU computing, GPU has become more and more important.

The status of GPU is becoming more and more important, but many consumers still seem to understand when it comes to graphics cards, especially for students who want to install the machine. In view of this, today we will comprehensively interpret the professional terms of graphics cards through easy-to-understand language, and easily create graphics card talents.

What is a graphics card?

The work of the graphics card is very complex, but its principles and components are easy to understand. In this article, let's first look at the basic components of graphics cards and their functions. In addition, we will examine the factors that work together to enable graphics cards to work quickly and efficiently.

The display card (videocard) is a necessary device for the system, which is responsible for processing the image data (data) sent by the CPU into a format that can be understood by the display (monitor), and then sending it to the display screen.

(screen) to form an image. It is the most important channel for us to get information from computers. Therefore, the display card and monitor is one of the most important parts of the computer. The image we see on the monitor is made up of many dots called "pixels". At the most commonly used resolution settings, the screen displays more than one million pixels, and the computer must decide what to do with each pixel in order to generate an image. To do this, it needs a "translator" who takes binary data from CPU and converts it into images visible to the human eye. Unless the computer's motherboard has a built-in graphics function, this conversion is done on the graphics card. As we all know, the computer is binary, that is, 0 and 1, but it is always missing to output 0 and 1 directly on the monitor, so there is a graphics card to convert these 0 and 1 into images to display.

The basic principle of graphics card

The main components of the video card are: motherboard connection device, monitor connection device, processor and memory. The working principle of different video cards is basically the same. CPU works with software applications to send information about images to the video card. The video card decides how to use the pixels on the screen to generate an image. It then sends this information to the monitor over a cable.

The evolution of graphics cards since IBM launched its first video card in 1981, graphics cards have been greatly improved. The first video card, called a monochrome display adapter (MDA), can only display green or white text on a black screen. Now, the minimum standard for new graphics cards is the Video Graphics Array (VGA), which can display 256 colors. With high-performance standards such as the Quantum expansion Map Matrix (QuantumExtendedGraphicsArray,QXGA), graphics cards can display millions of colors at resolutions up to 2040x1536 pixels.

Generating images from binary data is a laborious process. In order to generate a 3D image, the graphics card first creates a wireframe with a straight line. It then rasterizes the image (fills in the remaining pixels). In addition, the graphics card also needs to add light and shade light, texture and color. For fast-paced games, the computer must perform this process about 60 times per second. If there is no video card to perform the necessary calculations, the computer will not be able to bear such a heavy workload.

The four main components of the graphics card

When completing the work, the graphics card mainly depends on the coordination of four parts to complete the work, the motherboard is connected to the device for transmitting data and power supply, and the processor is used to decide how to deal with each pixel on the screen. the memory is used to store information about each pixel and temporarily store the completed image, and the monitor is connected to the device so that we can see the final result.

Processor and memory

Like the motherboard, the graphics card is a printed circuit board with a processor and RAM. In addition, it has an input / output system (BIOS) chip that stores the settings of the graphics card and performs diagnostics on memory, input, and output at startup. The processor of the graphics card is called the graphics processing unit (GPU), which is similar to the CPU of a computer. However, GPU is designed to perform complex mathematical and geometric calculations that are necessary for graphic rendering. Some of the fastest GPU have more transistors than the average CPU. GPU generates a lot of heat, so there is usually a radiator or fan installed above it.

In addition to its processing power, GPU uses special programming to help it analyze and use data. The vast majority of GPU on the market is produced by AMD and NV, and both companies have developed their own GPU performance enhancements. To improve image quality, these processors use panoramic anti-aliasing technology, which makes the edges of 3D objects smooth and anisotropic filtering, which makes the image look brighter.

When GPU generates an image, it needs a place to store information and completed images. This is the purpose of the graphics card RAM, which stores data about each pixel, the color of each pixel, and its location on the screen. Part of the RAM can also act as a frame buffer, which means that it will save the finished images until they are displayed. In general, the graphics card RAM runs at a very high speed and has a dual-port design, which means that the system can read and write to it at the same time.

The RAM is directly connected to the digital-to-analog converter, or DAC. This converter, also known as RAMDAC, is used to convert images into analog signals that can be used by monitors. Some graphics cards have multiple RAMDAC, which can improve performance and support multiple monitors.

Video card input and output

ADC Connector Apple has made monitors using the patented product AppleDisplayConnector (ADC). Although these monitors are still in use, Apple's new monitors have been switched to DVI-connected devices. The video card is connected to the computer motherboard through the motherboard to power the video card and enable it to communicate with the CPU. For higher-end graphics cards, the power provided by the motherboard is often insufficient, so the graphics card is also directly connected to the power supply of the computer.

The connection between the graphics card and the motherboard is usually realized by means of three interfaces: peripheral component interconnection (PCI), advanced graphics port (AGP) and PCIExpress (PCIe). Among these three interfaces, PCIExpress is the latest interface, which can provide the fastest transmission rate between the video card and the motherboard. In addition, PCIe supports the use of two graphics cards in one computer.

Most people use only one of the two monitors they have to connect to the device. Users who need to use two monitors can buy a graphics card with a dual-head output function, which can split the picture and display it on two screens. In theory, if the computer is equipped with two graphics cards with dual-head output and providing a PCIe interface, it can support four monitors. In addition to connecting devices for motherboards and monitors, some graphics cards also have connecting devices for the following purposes: TV display: TV output or S-Video, analog camera: ViVo (video input / video output, digital camera: hot wire or USB some graphics cards come with TV tuners.

Factors affecting the speed and efficiency of graphics cards

DirectX and OpenGLDirectX and OpenGL are application programming interfaces, referred to as API. API provides instructions for complex tasks, such as 3D rendering, to help hardware and software communicate more efficiently. Developers optimize games that use a lot of graphics for specific API. This is why the latest games usually require an updated version of DirectX or OpenGL to run correctly.

API is different from the driver. A driver is a program that enables hardware to communicate with the operating system of a computer. But like the updated version of API, the updated version of the device driver can help the program run correctly.

How to measure the quality of the graphics card?

The top-level graphics card is easy to identify and should have a lot of memory and a fast processor. In addition, it is usually the most interesting compared to any other part to be installed in the computer chassis. Many high-performance graphics cards claim to be required or directly equipped with exaggerated fans or radiators.

But the functions provided by high-end graphics cards exceed the real needs of most people. For users who mainly use computers to send and receive e-mail, word processing or surfing the Internet, the motherboard with integrated graphics cards can provide all the necessary graphics functions. For most users who occasionally play games, the mid-range graphics card is enough to meet the needs. Only game fans and users who need to do a lot of 3D graphics work need high-end graphics cards.

A good overall measure of graphics card performance is its frame rate, which is measured in frames per second (FPS). The frame rate indicates how many complete images the video card can display per second. The processing power of the human eye is about 25 frames per second, while fast-moving games require at least the frame rate of 60FPS to provide smooth animation and scrolling. Factors that affect the frame rate include the number of triangles or vertices generated per second. The 3D image is made up of triangles or polygons. This metric shows how quickly GPU can calculate the entire polygon or define the vertices of the polygon. In general, it shows how fast the graphics card can generate wireframe images.

Pixel fill rate: this indicator shows how many pixels GPU can handle in a second, which shows how fast the video card can rasterize the image. The hardware of the video card has a direct impact on its speed. The following are the hardware performance metrics that have the greatest impact on the video card speed and their units of measurement: GPU clock speed (MHz), memory bus capacity (bits), amount of memory available (MB), memory clock rate (MHz) memory bandwidth (GB/s), RAMDAC speed (MHz).

The CPU of the computer and the motherboard also have a certain impact on the speed of the graphics card, because the very fast graphics card can not make up for the lack of the motherboard's ability to transmit data quickly. Similarly, the connection between the video card and the motherboard and the speed at which it gets instructions from the CPU will affect its performance.

Overclocking some users choose to manually set the clock speed of their video card to a higher rate in order to improve the performance of the video card, which is called overclocking. People usually choose to overclock the memory of the video card because overclocking the GPU can lead to overheating. Although overclocking can achieve better performance, it can also invalidate the manufacturer's warranty.

Explanation of terms for main parameters of video card:

Display chip

Also known as pattern processor-GPU, it plays the same role in graphics cards as CPU does in computers. A more direct analogy is the role of the brain in the human body. GPU makes the graphics card less dependent on CPU and does part of the original CPU work, especially in 3D graphics processing. The core technologies used in GPU include hardware Thuml (geometric transformation and lighting processing), cubic environment material mapping and vertex blending, texture compression and bump mapping, dual texture four-pixel 256bit rendering engine and so on. Hardware Thuml technology can be said to be the symbol of GPU. The production of GPU is mainly produced by nVidia and ATI.

Development code

The so-called development code is the corresponding basic code given by the display chip manufacturer to a series of display chips in order to facilitate the management of the design, production and sales of display chips and the unity of driver architecture. The function of the development code is to reduce the cost of the display chip manufacturer, enrich the product line and unify the driver.

In general, display chip manufacturers can use a basic development code and then derive a series of display chips by controlling the number of rendering pipelines, the number of vertex shading units, the type of video memory, the width of the display memory bit, the core and memory frequency, and the supported technical characteristics. to meet different performance, price, market and other different positioning. In addition, the high-end display chip products with some defects in the manufacturing process can be sold into fully qualified corresponding low-end display chip products by shielding pipelines and other methods, so as to greatly reduce the difficulty and cost of design and manufacturing. enrich your product line. Display chips with the same development code can use the same driver, which provides convenience for display chip manufacturers to write drivers and consumers to use graphics cards.

Manufacturing process

Manufacturing process refers to the production of GPU in the process, to process a variety of circuits and electronic components, manufacturing wires to connect the various components. Usually the precision of its production is expressed in terms of nm (nanometer) (1mm=1000000nm). The higher the precision is, the more advanced the production process is. More electronic components can be made in the same material, and the thinner the connection line is, the more integrated the chip is, and the lower the power consumption of the chip is.

The development and progress of microelectronic technology mainly depends on the continuous improvement of process technology, so that the feature size of the device continues to shrink, so that the degree of integration is continuously improved, the power consumption is reduced, and the performance of the device is improved. Since 1995, the chip manufacturing process has gone from 0.5 micron, 0.35 micron, 0.25 micron, 0.18 micron, 0.15 micron, 0.13 micron, 0.09 micron, to the mainstream 65 nm, 55 nm, 40 nm.

Core frequency

The core frequency of the graphics card refers to the working frequency of the display core, and its working frequency can reflect the performance of the display core to a certain extent. however, the performance of the graphics card is determined by the core frequency, stream processor unit, video memory frequency, explicit memory bit width and so on, so when the display core is different, the high core frequency does not mean that the performance of the video card is strong. For example, the core frequency of GTS250 has reached

750MHz is higher than GTX260+ 's 576MHz, but GTX260+ is definitely better than GTS250 in performance. In the same level of chip, the performance of the chip with high core frequency is stronger, and improving the core frequency is one of the ways to overclock the graphics card.

Video card BIOS

Video card BIOS

It is mainly used to store the control program between the display chip and the driver, as well as the model, specification, manufacturer and factory time of the display card. When the computer is turned on, the information is fed back to the screen by displaying a control program in the BIOS. Early display BIOS is solidified in ROM and can not be modified, while most display cards use large-capacity EPROM, that is, the so-called Flash BIOS, which can be rewritten or upgraded through a special program.

Video memory

Displays the abbreviation of memory. As the name implies, its main function is to temporarily store the data to be processed by the display chip and the processed data. The stronger the performance of the graphics core, the more video memory is needed. In the past, the video memory is mainly SDR, the capacity is not large. Most of the graphics cards on the market use GDDR3 video memory, and now the latest graphics cards use GDDR4 or GDDR5 video memory with better performance. Video memory is mainly provided by traditional memory manufacturers, such as Samsung, Hyundai, Kingston and so on. The main type of video memory used on the video card is SDR.

DDR SDRAM,DDR SGRAM 、 DDR2 、 DDR3 、 DDR4 、 DDR5 .

DDR SGRAM

It is a product improved by graphics card manufacturers from synchronous dynamic random access memory (SDRAM) to meet the needs of drafters and to enhance the efficiency of graphics access processing and drawing control. SGRAM allows data in memory to be modified or accessed individually in blocks (Blocks). It can work synchronously with the central processing unit (CPU). It can reduce the number of memory reads and increase the efficiency of the drawing controller. Although it has good stability and good performance, its overclocking performance is very poor.

Memory interface

The bit width of explicit memory is the number of bits that can be transmitted in a clock cycle. The larger the number of bits, the greater the amount of data that can be transmitted instantly, which is one of the important parameters of video memory. In 2009, there are 64-bit, 128-bit, 256-bit and 512-bit video memory widths on the market. The 64-bit, 128-bit and 256-bit video cards that people are used to call refer to their corresponding video memory bit widths. The higher the video memory bit width, the better the performance, the higher the price, so the 512-bit wide video memory is more used in high-end graphics cards, while the mainstream graphics cards basically use 128-bit and 256-bit video memory.

Video memory bandwidth = video memory frequency X display memory bit width / 8, in the case of the same video memory frequency, the video memory bit width will determine the size of the video memory bandwidth. The video memory of the video card is composed of a piece of video memory chips, and the total bit width of the video memory is also composed of the bit width of the video memory particles. Apparent memory bit width = apparent memory particle width × number of apparent memory particles. Graphics memory particles are all with the relevant manufacturer's memory number, you can go to the Internet to find its number, you can understand its bit width, and then multiplied by the number of graphics memory particles, you can get the bit width of the card.

Video memory speed

The speed of video memory is generally measured in ns (nanoseconds). Common video memory speeds are 1.2ns, 1.0ns, 0.8ns, etc., and the smaller the speed, the faster and better. The formula for calculating the theoretical working frequency of video memory is as follows: equivalent working frequency (MHz) = 1000 / (video memory speed × n) (n varies with different types of video memory, if it is GDDR3 memory, then nasty 2 / GDDR5 memory.

Memory frequency

To some extent, the video memory frequency reflects the speed of the video memory. In MHz (megahertz), the video memory frequency varies with the type and performance of the video memory. DDR SDRAM video memory can provide a higher video memory frequency, so it is the most widely used type of video memory. DDR is used in most of the middle and low-end graphics cards and high-end graphics cards.

SDRAM, which can provide video memory frequency is also very different, mainly 400MHz, 500MHz, 600MHz, 650MHz and so on, high-end products also have 800MHz or 900MHz, or even higher.

Stream processor unit

Before the DX10 graphics card came out, there was no such thing as a "stream processor". The interior of GPU consists of "pipelines", which are divided into pixel pipelines and vertex pipelines, and their number is fixed. To put it simply, the vertex pipeline is mainly responsible for 3D modeling, and the pixel pipeline is responsible for 3D rendering. Because the number of them is fixed, there is a problem that when a game scene requires a lot of 3D modeling and does not need too many pixels, it will result in a shortage of vertex pipeline resources and a large number of pixel lines idle. Of course, there is another situation on the contrary.

Under such circumstances, people put forward the "unified rendering architecture" for the first time in the DX10 era. The graphics card canceled the traditional "pixel pipeline" and "vertex pipeline" and changed it into a stream processor unit, which can perform both vertex operations and pixel operations, so that in different scenarios, the graphics card can dynamically allocate the number of stream processors for fixed-point operations and pixel operations, so as to make full use of resources. Now, the number of stream processors has become a very important indicator to determine the performance of graphics cards. Nvidia and AMD-ATI are also constantly increasing the number of stream processors of graphics cards to make the performance of graphics cards grow by leaps and bounds. It is worth mentioning that the GPU architecture of N card and A card is not the same, and the allocation of stream processor number is not the same.

Double card technology

SLI and CrossFire are the working group modes of dual-card or multi-card interconnection of Nvidia and ATI respectively.

Its essence is similar. But the name is different, SLI Scan Line Interlace (scan line interleaving) technology is applied by 3dfx on Voodoo. It physically connects two Voodoo cards with SLI lines. When working, one Voodoo card is responsible for rendering odd-numbered line scanning of the screen, and the other is responsible for rendering even-numbered line scanning, thus achieving the "double" performance of "connecting" two video cards together. SLI Chinese name Speed Force, in 2009, the working mode of SLI was different from that of the early Voodoo, and it was changed to screen partition rendering.

DirectX

DirectX is not a simple graphic API, it is a widely used API developed by Microsoft, which includes Direct Graphics (Direct 3D+Direct Draw), Direct Input, Direct Play and Direct.

Sound, Direct Show, Direct Setup, Direct Media Objects and other components, it provides a complete set of multimedia interface scheme. It's just that its excellent performance in 3D graphics makes other aspects of it look dim. At the beginning of its development, DirectX is to make up for the lack of graphics and sound processing capacity of Windows 3.1. it has developed into an interface that has a decisive impact on all aspects of the whole multimedia system, and the latest version is DirectX 11.

About how the GPU parameters are shared here, I hope the above content can be of some help to you, can learn more knowledge. If you think the article is good, you can share it for more people to see.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.