computer told help answer cards

There are many models of graphics cards are generally alphanumeric with expert can be very simple and clear pointers about how the meaning of alphanumeric letters and numbers by type and graphics on the show good or bad Thanks a # -0 – # nVidia's: GS not the know, only know that GS is the main product of the GT series is a higher level than the GS is the high end of the series, a little taste of GT cars in the high-end high-end GTX is generally understood as GT eXtreme LE is the simplified version, he should be Limited Editon TC, TurboCache shorthand, nVidia, a technique that will allow graphics cards alone can be shared as video memory and a memory is Ultra, before the GF7 series represents the most high-end, but the naming of the most high-end 7 series turned into GTX ATI's: SE, Special Editon of shorthand, like nVidia's LE HM, HyperMemory, and a reason above TC Pro, the X100 series before the Pro version represents a higher level than normal version, but after the X100 series and represents the mainstream version, such as the X700Pro, X1600Pro XT, a higher level than the Pro version, the X100 series is the most high-end before, X100 series and is second after the high-end XT PE, XT PE on behalf of XT Premium Edition, the XT Platinum Edition, X100 series appeared, the most high-end models XTX, X1000 series named after the release of the new rules, and the GTX a reason, XT eXtreme XL, high-end series in the lower end models concrete can be understood as eXtreme Limited GTO, even lower level than the XL model some of the latter two models are mixed together, such as GT, XT, in the nVidia, GT represents the higher end of the first after a GTX Second, strong, and in the ATi in, GT is a simplified version of XT. XT represents the nVidia high-end series in the most simplified version, and in ATI, XTX or XT PE is the second after the strong did not say what the best with this statement, is to belong to the same level of CPU and graphics card What? X1600PRO or 7600GS suffix Xiangjie: LE: NVIDA graphics card model with the suffix. Full name is “Limited Edition” (limited edition) represents the series of low-end products, currently used only GeForce FX 5700. SE: ATI graphics card model with the suffix. Full name of the “Special Edition” (special edition), the same low-end products on behalf of the series. SE suffix cards are usually only a 64bit memory interface, such as the 9200SE, 9550SE, 9600SE, 9800SE (this model with 128 bit and 256 bit), X300SE so. Or reduce the number of lines of pixels (such as the 9800SE). ZT: NVIDA additional models, there are only a GeForce FX 5900. ZT represents less than the XT's market position, so that the series of low-end products. XT: the most confusing subject models. ATI and NVIDA cards are in use. In the ATI, the representative of the most high-end series of products, such as 9600XT, 9800XT, X800XT, X600XT so. The NVIDA will be used as a low-end models, such as the 5600XT, 5900XT so. GT / Pro: respectively, NVIDA and ATI graphics models of the company as the middle of the suffix. Representative products 6800GT, 9600Pro, X800 Pro, the only exception is the new GeForce 6600 GT graphics card, which is the series for the time being the most high-end graphics card. Ultra: NVIDA graphics card in the top of the card, and ATI in the “XT” similar. Representative products are 6800 Ultra, 5700Ultra NV's GT XT expensive than is normal, because the XT is the most low-end models .. NV # -0 – # Introduction of new technologies now emerging graphics card, the parameters of weight and more, new to the hardware Friends often want to know a card good or bad but not from the looks, tell us about the graphics in this special parameters of the relevant property, what good or bad, you can decide it. Appropriate, is the best. The main composition graphics (most parameters) 1, display chip (type, version level, development code, manufacturing process, the core frequency) 2, memory (type, bit width, capacity, package type, speed, frequency) 3, Technology ( pixel rendering pipelines, the number of vertex shaders, 3D API, RAMDAC frequency and support the MAX resolution) 4, PCB board (PCB layers, graphics interface, output interface, heat sink) 5, Brand 1, showing the chip graphics chips, but also said graphics processor – GPU, it's role in the graphics, just as the role of CPU in the same computer. More direct analogy is that the brain's role in the human body. First a brief overview of common manufacturing graphics chips vendors: Intel, ATI, nVidia, VIA (S3), SIS, Matrox, 3D Labs. Intel, VIA (S3), SIS produces integrated chips; ATI, nVidia chip based independent, is the mainstream market, but AMD ATi has now been acquired, will continue after a separate graphics chip, hard to say; Matrox, 3D Labs is primarily for professional graphics market. ATI and nVidia as the basic occupied the mainstream graphics market, mainly targeted at the following two products do introduction. The company's main brand models ATi Radeon (radium) series, the models by the early its the Radeon Xpress 200 to the Radeon (X300, X550, X600, X700, X800, X850) to the recent Radeon (X1300, X1600, X1800, X1900, X1950 ) performance order from low to high. The company's main brand nVIDIA GeForce series, the models by the early its the GeForce 256, GeForce2 (100/200/400), GeForce3 (200/500), GeForce4 (420/440/460/4000/4200/4400/4600/4800 ) to the GeForce FX (5200/5500/5600/5700/5800/5900/5950), GeForce (6100/6150/6200/6400/6500/6600/6800 /) and then to near its the GeForce (7300/7600/7800 / 7900/7950) performance order from low to high. In addition to the standard version of the version level but also some special edition, special edition will normally be the model behind the Standard Edition plus a suffix, common are: ATi: SE (Simplify Edition Starter Edition) is usually only 64bit memory interface, or pixel reduction in the number line. Pro (Professional Edition Professional Edition) high-frequency version of the general quantity than the standard version in the pipelines / vertex number of the frequency of these aspects have to be slightly higher. XT (eXTreme high-end version) is a series of high-end ATi and nVIDIA for low-end models. XT PE (eXTreme Premium Edition XT Platinum Edition) high-end models. XL (eXtreme Limited high-end series in the lower end models) ATI's new R430 in the high-frequency version of XTX (XT eXtreme high-end version) X1000 series named after the release of the new rules. CE (Crossfire Edition Crossfire Edition) cross-fire. VIVO (VIDEO IN and VIDEO OUT) means the graphics card also has two video input and video capture capabilities. HM (Hyper Memory) memory cards can take up nVIDIA: ZT XT based on the frequency down to lower prices again. Down XT version, but in the most high-end ATi said. LE (Lower Edition low-end version) and basically the same as XT, ATi also used. MX cheap version of the general public category. GTS / GS low frequency version. GE slightly stronger than the GS point, in fact, over the frequency of the GS. GT high-frequency version. A grade higher than GS because GT did not reduce the pipelines and vertex units. GTO slightly stronger than the GT point, a little car in the GTO flavor. Ultra before the GF7 series represents the most high-end, but the most high-end 7 series to the naming of the GTX. GTX (GT eXtreme) enhanced version, reduced frequency or reduced water pipes became the GT, and then continue to shrink a GS version. GT2 dual GPU graphics card. TI (Titanium Ti) is the representative of the general high-end version of nVidia. Go multi-language mobile platform. TC (Turbo Cache) can take up a memory card is the codename for a so-called development of graphics chips, code-named manufacturer of graphics chips in order to facilitate the design, production, sales, management and unified driver architecture to a series of graphics chips give the corresponding basic code. Development of graphics chips, code-named role is to reduce the cost of the manufacturer, rich product line and to achieve the unity of the driver. In general, the display chip manufacturers can use a basic development code and then by controlling the number of rendering pipelines, the number of vertex shader units, memory type, memory interface, the core and memory frequencies, technical characteristics of the support to derivative terms a series of graphics chips to meet the different performance, price, market and other various Dingwei include putting the manufacture Guocheng Gaoduan has some flaws in graphics chips and other products handled by pipeline screen as Wanquanhege of Xiang Ying low end of graphics chips products sold, thus substantially reducing design and manufacturing difficulties and costs, enrich their own product line. The same chip, code-named the show can use the same driver, this chip makers prepared to show drivers and graphics cards offer consumers the convenience. The same chip, code-named the show as well as to support rendering architecture is essentially the same technical characteristics, and also used the same process, so the development code is to judge video card performance and grades of the important parameters. Number of different versions of the same type can be a code, for example: GeForce (X700, X700 Pro, X700 XT) code is RV410; and Radeon (X1900, X1900XT, X1900XTX) code is R580, etc., but there are other circumstances, such as : GeForce (7300 LE, 7300 GS), code-named G72; the GeForce (7300 GT, 7600 GS, 7600 GT) are the G73 and other code. Manufacturing Process Manufacturing Process refers to the GPU in the production process, to conduct processing of various circuits and electronic components, manufacturing wires connect the components. Usually the accuracy of their production to um (micron) to that future to the nm (nanometer) trend (1mm = 1000um 1um = 1000nm), the higher the precision, more advanced production technology. In the same material can create more electronic components, cables are thinner, improving chip integration, power consumption is smaller. Micron manufacturing process is the IC circuit within the circuit and the distance between. Manufacturing technology trend is to the direction of higher density. The higher density IC circuit design, means that in the same area the size of IC, you can have a higher density, more complex circuit design. Microelectronics technology development and progress, mainly through the continuous improvement of process technology, making the device feature size shrinking to continuously improve the integration, power consumption, device performance is improved. Chip manufacturing processes after 1995, from 0.5 micron, 0.35 micron, 0.25 micron, 0.18 micron, 0.15 micron, 0.13 micron, and then to present the mainstream 90-nanometer (0.09 nm), 65nm and so on. Graphics core frequency core frequency is shown the core operating frequency, the frequency of its work to some extent reflects that show the performance of the core, but the graphics performance is the core frequency, memory, pixel pipelines, pixel fill rate and so many determined by the situation, so different in the core of the case show that the core does not mean that the high frequency of strong graphics performance. For example 9600PRO core frequency reached 400MHz, the 380MHz higher than the 9800PRO, but definitely better than the performance 9800PRO 9600PRO. Chips in the same level, while high frequency performance of the core of some stronger, improve the core frequency is one way to overclocking graphics cards. Mainstream ATI graphics chips and NVIDIA only two, both are core to a third party to provide display manufacturers, display the same core, some manufacturers will display their products appropriately increasing the core frequency, the core of its work in the above display a fixed frequency to achieve higher performance. 2, memory type used in the current market, type of main memory SDRAM, DDR SDRAM, DDR SGRAM of three. SDRAM particles is mainly applied in the low-end graphics card, the frequency of not more than 200MHz, the price and performance than DDR is no advantage, therefore gradually being replaced by DDR. DDR SDRAM is the abbreviation for Double Data Rate SDRAM (double data rate), it can provide a higher operating frequency, to bring superior data processing performance. DDR SGRAM is a graphics card specifically for the demands of manufacturers, in order to enhance access to graphics processing and graphics control efficiency, from the synchronous dynamic random access memory (SDRAM) obtained by the improved products. SGRAM allows the box (Blocks) to alter or access to individual units of memory in the data, it can with the central processing unit (CPU) synchronization can reduce the number of memory reads, increasing the efficiency of graphics controller, although it stability Yes, but performance is also very good, but its poor overclocking performance. The mainstream of the market is DDR2 and DDR3,. Bit wide memory interface is the memory can transmit data in a single clock cycle the number of bits, the greater is the median amount of data can instantly bigger, this is one of the important parameters of video memory. Currently on the market are 64-bit memory interface, 128-bit, 256 bit and 512 some people used the name of the 64-bit graphics, 128 graphics and 256 graphics cards is that the corresponding memory interface. The higher the memory interface, the better the performance the higher the price, so 512-bit wide memory more high-end graphics cards, and basically all mainstream video card with 128 and 256-bit memory. = Memory memory bandwidth memory interface frequency X / 8, in the case of memory frequency rather, memory interface will determine the size of memory bandwidth. For example: the same memory frequency is 500MHz, 128-bit and 256-bit memory, so Talia's memory bandwidth will be as follows: 128 = 500MHz * 128 / 8 = 8GB / s, and 256 = 500MHz * 256 / 8 = 16GB / s, is 2 times 128, memory interface shows the importance of data in memory. The memory card is a piece of the memory chip components, the total bit wide memory is also formed by the bit width memory particles. Particle-bit wide memory interface = memory × memory particles. Particles are associated with memory memory manufacturers number, you can go online to find their number, we can understand the bit width, multiplied by the number of memory particles, can be a bit wide graphics cards. Capacity of this relatively well understood, the greater the capacity, the more things exist, of course, the better. Current mainstream memory capacity, 64MB, 128MB, 256MB, 512MB and so on. Package Type Package main memory: TSOP (Thin Small Out-Line Package) thin small outline package QFP (Quad Flat Package) small quad flat package MicroBGA (Micro Ball Grid Array) micro-BGA, also known as FBGA ( Fine-pitch Ball Grid Array) is basically the current mainstream graphic cards and MBGA with TSOP package, which went mostly TSOP package. speed memory speed generally ns (nanoseconds) for the unit. Common memory speed 7ns, 6ns, 5.5ns, 5ns, 4ns, 3.6ns, 2.8ns, 2.2ns, 1.1ns and so on, the less said the faster \ \ better. The theory of memory frequency formula is: Rated operating frequency (MHz) = 1000 / memory speed × n be (n for memory types differ, if it is SDRAM memory, then n = 1; DDR memory is n = 2; DDRII memory is n = 4). Memory clock frequency to some extent with the memory speed of response to MHz (megahertz) for the unit. With the memory of the type of memory frequency, performance varies: SDRAM memory generally work at lower frequencies, the general is the 133MHz and 166MHz, the frequency of such cards are already unable to meet the demand. DDR SDRAM memory is able to provide higher memory frequencies, so is the most widely used memory type, present both in low-end graphics cards, or most of the high-end graphics cards use DDR SDRAM, it can provide significant deposit very different frequency, mainly 400MHz, 500MHz, 600MHz, 650MHz and so on, high-end products have 800MHz or 900MHz, and even higher. Memory frequency and memory clock cycles are related, reciprocal relations between the two as that is, frequency = 1 memory / memory clock cycles. If SDRAM memory, its clock cycle 6ns, then its memory frequency on the 1/6ns = 166 MHz; while for DDR SDRAM, the clock period of 6ns, then its memory frequency on the 1/6ns = 166 MHz, but to understand is that this is the actual frequency of DDR SDRAM, not DDR memory we usually say the frequency. Because DDR clock rise and fall of all of the data transmission, one of its two data transfer cycles, the equivalent of twice the SDRAM frequency. Customary to refer to the DDR frequency is the equivalent frequency, is superior in its actual working frequency of 2, we get the equivalent frequency. So 6ns of DDR memory, the memory frequency 1/6ns * 2 = 333 MHz. But to understand is that graphics card manufacturers, the manufacturers set the memory frequency of the actual work, but the actual memory frequency does not necessarily mean that the maximum frequency. Such situation is now more common, such as the maximum memory to work in 650 MHz, while the manufacture graphics cards operating frequency is set to 550 MHz, then there are some overclocking memory on spatial. This is the usual method of the current manufacturers, graphics card with overclocking as a selling point. 3, Technical pixel pipelines rendering pipeline, also known as rendering pipeline, is to show graphics chip deal with separate parallel signal processing unit. Rendering pipeline to some extent can be compared to the factory inside the common variety of production lines, the factory's production line is to increase production capacity and efficiency, and improve the graphics rendering pipeline is the ability to work and efficiency. The number of rendering pipelines is generally the number of pixel lines × per pipeline, said the number of texture units. For example, GeForce 6800Ultra rendering pipeline is 16 × 1, says that its has a 16 pixel pipeline, each pipeline has a texture unit; GeForce4 MX440 rendering pipeline is the 2 × 2, says that its has two pixel rendering pipeline, Each pipeline with two texture units, etc., that way the rest and so on. Rendering pipeline is to determine the number of display chip performance and grades of one of the most important parameters, the same graphics core frequency, the more pipelines means greater pixel fill rate and texture fill rate, from the graphics card volume rendering pipeline can largely determine the grade level of graphics performance. But graphics performance was not just depend on the number of rendering pipelines, and also depends on the graphics core architecture, rendering pipeline execution efficiency, the number of vertex shader units and the graphics core frequency and memory frequency, etc.,. Generally speaking, the core structure in the same display, the rendering pipeline more means higher performance, such as 16 × 1 structure of the GeForce 6800GT performance better than 12 × 1 structure of the GeForce 6800, like in a factory using the same technology, two production lines of production capacity and efficiency of better than a production line as; in different display core structure, the number of rendering pipelines and more on does not mean better performance, such as 4 × 2 structure of the GeForce2 GTS its performance as good as 2 × 2 framework GeForce4 MX440, as the factory's use of advanced technology, a line of productivity and efficiency but also better than the old technology, using only two production lines did. The number of vertex shader vertex shader (Vertex Shader), also known as vertex shielding device, according to official specifications, to increase vertex shader is a variety of special effects in 3D games shadow of the processing unit, vertex shader, programmable features by allowing developers to add new kinds of software instructions to adjust the effects, each vertex will be a variety of data variables have clearly defined, including at least each of the vertices x, y, z coordinates of each vertex point may package Functional data are color, original path, material, light characteristics. More the number of vertex shader faster. 3D APIAPI is the acronym for Application Programming Interface, is the application program interface means, and 3D API refers to the graphics interface directly with the application. 3D API designed to allow programmers to call its API 3D software as long as the procedures, so that API and hardware driver automatically communicate a strong start the 3D-chip 3D graphics capabilities, which greatly improved the 3D process design efficiency. If there is no 3D API in the development process, the programmer must understand all of the graphics capabilities, to prepare a complete match with the graphics process play out all the graphics performance. But also has a 3D API that directly interface card and software, programmers only need to write code line interface, you can give full play to graphics hardware do not go to understand the specific properties and parameters, thus greatly simplifying the program development efficiency . Similarly, the display chip makers to design their own based on standard hardware products to meet the API calls to optimize hardware resources, for better performance. With the 3D API, can achieve different manufacturers of hardware and software compatible with the widest range. For example, in the best embodiment 3D API's games, game design, to design O'clock do not have to consider the details of specific models Xianka the Texing, Er Zhishianzhao 3D API's interface to the standard to develop game, when the game when Ze Yun Hang directly through the 3D API to call graphics hardware resources. Currently the main application of PC 3D API are: DirectX and OpenGL. RAMDAC RAMDAC frequency and support the maximum resolution is Random Access Memory Digital / Analog Convertor acronym, that is, random access memory digital ~ analog converter. RAMDAC memory function is to convert the digital signal display can show up in the analog signal, the conversion rate in MHz indicated. Computer data processing is actually done in the digital process will be things, all things will be processed into two numbers 0 and 1, and then ongoing cumulative. Graphics accelerator card is by the 0 and 1 on the pixels of each color, depth, brightness and other processing. Xianka signals are generated are based on figures expressed, but all the CRT monitors are the way to simulate the work of, digital signal can not Pishi Bie, which must have appropriate equipment will be digital signal converted to analog signals. The RAMDAC is the graphics card will convert digital signals to analog equipment. MHz RAMDAC conversion rate to that which determines the refresh rate of the high and low (and display the “bandwidth” meaning similar). The higher the speed of its work, the more wide-band, high-resolution picture quality better when. The value determined is sufficient memory, the graphics cards support the highest resolution and refresh rate. If you are under 1024 × 768 resolution to 85Hz resolution, RAMDAC rate of at least 1024 × 768 × 85 × 1.344 (conversion factor) ÷ 106 ≈ 90MHz. Current mainstream graphics card RAMDAC can reach 350MHz and 400MHz, and more than sufficient to meet the most current display resolution and refresh rate can provide. 4, PCB board PCB is the acronym for Printed Circuit Block, also known as printed circuit boards. Is the card body (green board), all components are on the graphics card PCB board, PCB board good or bad, therefore, directly determines the electrical properties of good and bad graphics and stability. PCB board layers are usually present in 4 layer, 6 layer, or 8 layers, in theory, fewer layers and more than a good, but only in the design of rational basis. Each layer of PCB is generally divided into signal layer (Signal), the power level (Power) or the ground floor (Ground). Each layer PCB version of the circuit are independent of each other. 4 layer PCB board, the signal level of the general distribution of the PCB and the bottom layer of the top layer, while the middle two are power and ground layer. 6 layer PCB, relatively speaking on the complex, the signal level usually found in 1,3,5-layer, and power planes have two layers. As for judging the merits of PCB, mainly to observe whether the clarity of some of the printed circuit, PCB is so smooth and no deformation. Graphics interface, common with PCI, AGP 2X/4X/8X (now obsolete), the latest is PCI-Express X16 interface, is now mainstream. Output interface is the most common main output interface: VGA (Video Graphics Array) video graphics array interface, the role is to convert a good analog signal output to the CRT or LCD monitors in DVI (Digital Visual Interface) digital video interface, interface, video signal No conversion, no signal attenuation or distortion, the future replacement for VGA interface. S-Video (Separate Video) S terminal, also known as two component video interface, generally use five-line connector, which is used to separate brightness and color output devices, the main function is to overcome the composite video output brightness with chroma interference. The quality of heat sink cooling device can also affect the operation of graphics stability, a common cooling device are: passive cooling: only just installed the aluminum or copper and other metal heat sink. Air Cooling: to install the heat sink fan, now most of this approach. Water cooling: liquid through the heat pipe connected to the GPU and the pump is generally used in high-top graphics. Many people think that the red color graphics better than the green, green, yellow tongue than good, graphics good or bad, and its color does not matter, and some manufacturers like to use red, some like to use green, which is produced entirely by business decisions. Some brand-name manufacturers, it is already formed a certain style. Therefore, the PCB color generally will not be much change. Brand competition in the current graphics card industry is increasingly fierce. The numerous kinds of brands, the following is my view that some of the more good sign for reference only, do not be too superstitious: Mai Chong (MATROX), 3Dlabs, sapphire (SAPPHIRE), Asustek (ASUS), Hon Hai (Foxconn), (PowerColor / Dataland), Leadtek (Leadtek), XFX (XFX), reflecting the public (Inno3D) Micro-Star International (MSI), Elsa (ELSA), rich color (FORSA), Palit (Palit), Jetway ( Jetway), Abit (Abit), Epox (EPOX), Biostar (Biostar), Gainward (Gainward), SPARKLE (SPARKLE), Ying Chi (GALAXY), day of Jan (GRANDMARS), superior-day color (SuperGrece ), MAXSUN (MAXSUN), ASL (ASL), Yeston (YESTON) # -0 – # to see it bubble net, which select the graphics ~ ~ first all prices and specifications are familiar with the ~ ~ to Here, also said the next one or two graphics chips do not know ~ ~ mainly on two companies, one is NVIDIA. One is the ATI. The following is the response, performance from the top, followed by decreasing. Performance with the row close to ~ ~ NVIDIA VS ATI GTX275 HD4890 GTX260 + HD4870 GTS250 HD4860 or HD48509800GT HD4750 4830, early HD3870 9600GT HD3850 9600GSO HD4670 9500GT HD4650 is more than the mainstream of mid-ranking in 2009, the development of fast graphics card, NVIDIA launched this year GTX480/470/460. All three are high-end products, prices around the turn 3888,2888,1500 ATI HD5000 series graphics card launched, is the most high-end HD5970, to more than 5000 RMB ~ ~ recommend ranked online to see if card (Baidu look, there are many) Some ranking and the other some of the access ~ ~ ~ know most about the performance graphics card, after the bubble net, Zhongguancun this site to see the price of each card, specifications ~~~~~ This is not a day or two will be able to ~ ~ part of the master card's performance ranking (300 yuan to 1,500 yuan) G210 <9500GT (≈ GT220) <HD4670 ≈ 9600GS0 <9600GT (GT240) <HD4830 ≈ 9800GT <GTS250 (a little stronger than the 9800GT, not much difference) ≈ HD4850 <GTX260 ≈ HD4870 ≈ HD5770 <GTX460 with the price of ATI graphics performance over the NVIDIA graphics card (mainly HD4000 series, HD5000 series, part of the cost is not high, HD5000 main support DX11, energy)

References : If your answer is quoted from elsewhere, please indicate the source # -0 – # I do not know very much, graphics card brand and model of the mainstream market, many major chip has two types of ATI, a kind of NViDIA, high school low grades of all, mainly to see the economic situation of your use and you type a search online, the price came out ~ one point the price of a sub-goods, good things certainly expensive, you know the landlord