the true religion womens Memory card on the main t

SDRAM, DDR SGRAM, DDR2, GDDR2, DDR3, GDDR3, GDDR4, GDDR5. DDR SDRAM is the abbreviation for Double Data Rate SDRAM (double data rate), it can provide a higher operating frequency, a superior data processing performance. DDR SGRAM is specifically for graphics cards manufacturers who demand access to treatment in order to enhance graphics and drawing control efficiency, from the synchronous dynamic random access memory (SDRAM) is obtained by the improved products. SGRAM allows the box (Blocks) to modify or access the individual units of data in memory, it can with the central processing unit (CPU) synchronization can reduce the number of memory reads, increasing the efficiency of the graphics controller, although it stability Yes, but performance is also very good, but its poor overclocking performance. The current mainstream GDDR3 and GDDR5. Memory Interface is a memory that can transmit data in a number of bits per clock cycle, the greater the number of bits that can transmit the same frequency the greater the amount of data. 2010 graphics card on the market are 128-bit memory interface, 192 bits, 256 bits of several. The memory bandwidth memory interface = Memory Frequency X / 8, which represents the memory of the data transmission speed. Considerable frequency in the case of memory, memory interface will determine the size of memory bandwidth. For example: the same memory frequency is 500MHz, 128-bit and 256-bit memory, the memory bandwidth they are: 128 = 500MHz * 128 / 8 = 8GB / s; and 256 = 500MHz * 256 / 8 = 16GB / s, is 2 times 128. The memory card is a memory chip, consisting of the block, the total bit wide memory is also a bit wide of particles formed by the memory. Memory Interface Width × memory = memory particles number of particles. Memory manufacturers particles are associated with the memory number, you can go online to find their number, we can understand the bit width, multiplied by the number of memory particles, can get a bit wide graphics cards. Other specifications of the same graphics card, the bigger the better the performance bit wide. The case of other parameters the same capacity the bigger the better, but more can not focus only on the memory card (a lot of the core js will be equipped with large memory and low performance as a selling point.) For example, the 9600GSO 384M 512M to much stronger than the 9600GSO, because the core and memory bandwidth are quite different. Select graphics card memory capacity is only one reference to the core and bandwidth and other factors more important, these decisions take precedence over the performance of graphics card memory capacity. But the necessary memory capacity is necessary, because in the case of high-resolution high anti-aliasing may appear insufficient memory. The current market graphics memory capacity ranging from 256MB-4GB. TSOP (Thin Small Out-Line Package) thin small outline package QFP (Quad Flat Package) small quad flat package MicroBGA (Micro Ball Grid Array) micro ball grid array package, also known as the FBGA (Fine-pitch Ball Grid Array) 2004 年the former is basically the mainstream graphics and MBGA TSOP package, TSOP package majority. However, due to nvidia's gf3, 4 lines appear, MBGA into the mainstream, mbga package can achieve faster memory speeds, far exceeding the limits of TSOP 400MHZ. Memory speed is generally in ns (nanoseconds) for the unit. Common memory speed 1.2ns, 1.0ns, 0.8ns and so on, said the smaller the faster the better. Theoretical work of video memory frequency calculation formula is: the equivalent frequency (MHz) = 1000 × n / (memory speed) (n for memory types differ, if it is GDDR3, the n = 2; GDDR5 memory the n = 4). Memory frequency response to a certain extent, the speed of the memory to MHz (megahertz) for the unit. High and low frequency of memory and memory types have a very strong relationship: SDRAM memory generally work at lower frequencies, this frequency has long been unable to meet the graphics needs. DDR SDRAM memory is able to provide higher memory frequencies, we now have adopted the basic graphics DDR SDRAM, the memory frequency can provide very different. Has now developed to GDDR5, the default is equivalent to the maximum operating frequency has reached 4800MHZ, but still a great potential to improve. Memory clock frequency and memory are related, both as the reciprocal of, that is, memory frequency (MHz) = 1 / memory clock cycle (NS) X1000. If the SDRAM memory, the clock period is 6ns, then its memory frequency to the 1/6ns = 166 MHz; and for DDR SDRAM, the clock period of 6ns, then its memory frequency to the 1/6ns = 166 MHz, but to understand is that this is the actual frequency of DDR SDRAM, not DDR memory frequency usually said. Because DDR clock rise and fall of all of the data transmission, a period twice the data transfer, the equivalent of twice the frequency of SDRAM. Customary to refer to the DDR frequency is the equivalent frequency, is superior in its actual working frequency of the equivalent frequency of 2. Therefore, the DDR 6ns 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 the maximum frequency, such cases are more common. But there are also the largest memory can not be the nominal operating frequency of stable jobs. DX10 graphics cards out in the past, and there is no GPU inside the Simply put, the vertex pipeline is responsible for 3D modeling, 3D rendering pixel pipelines responsible. Since their number is fixed, there is a problem which, when a game requires a lot of 3D modeling of scenes without too much pixel processing, vertex pipelines will cause strain on the resources and a large number of pixel pipelines idle, of course, there are Another opposite situation. This will cause some of the resources is not enough and some other waste of resources. In such circumstances, it is the first time in the DX10 era of the pixel operation can be, so that different scenarios, graphics can be dynamically allocated to vertex operations and pixel operations for the stream processors to achieve full utilization of resources. Now, the number of stream processors, a decision on how much graphics performance has become a very important level indicator, Nvidia and AMD-ATI graphics cards are constantly increasing the number of stream processors so that graphics performance to jump in, such as AMD -ATI HD3870 graphics card with 320 stream processors, HD4870 reaches 800, HD5870 is to reach 1600! It is worth mentioning that, N A card-GPU card and not the same as the distribution for the number of stream processors are not the same. N card for each stream processor unit contains only one stream processors, while the A card, which is equivalent to each stream processor units with five stream processors, such as HD4850 ​​800 stream processors, though, in fact, only about 160 stream processor units, the other A stream processor card, the same frequency and core frequency, which is why the 9800GTX + is only 128 stream processors, the performance with the HD4850 ​​is quite a (N card stream processor core frequency, the frequency is about 2.16 times). Application Programming Interface API is an abbreviation, the meaning is the application program interface, 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 the API and hardware drivers automatically communicate a strong start the 3D-chip 3D graphics processing capabilities, which greatly improved the 3D program design efficiency. If there is no 3D API, the programmer in the development of procedures must understand the characteristics of all the cards in order to write an exact match with the graphics program, to play all of the graphics performance. And with this video card with 3D API software interfaces directly, the programmer need only comply with the interface of the program code written to take full advantage of graphics performance, do not have to understand the specific properties and parameters of the hardware, thus greatly simplifies program development efficiency. Similarly, the graphics chip manufacturers 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 be realized by different manufacturers of hardware and software compatible with the widest range. For example, in most of the game reflect the 3D API, the game designers design, do not have to consider the specific characteristics of a graphics cards, but only in accordance with the interface standard 3D API to develop the game, when the game is running directly through the 3D API to call graphics hardware resources. The main application of personal computers with 3D API: DirectX and OpenGL. RAMDAC is a Random Access Memory Digital / Analog Convertor the acronym for random access memory digital-analog converter ~. RAMDAC memory function is to convert the digital signal out of your monitor can display analog signal, the conversion rate in MHz said. Computer data processing is actually done things in the digital process will be, all things will be treated as two numbers 0 and 1, and then ongoing cumulative. By the graphics card is 0 and 1 for each pixel for color, depth, brightness, and other treatment. Graphics signals are generated digitally represented, but all the CRT monitors are based on simulation work, the digital signal can not be identified, which must have the appropriate equipment, the digital signal converted to analog signals. The RAMDAC is the graphics card will be the digital signal converted to analog signal equipment. MHz RAMDAC conversion rate to that which determines the level of refresh rate (and the display of the The higher the speed of its work, the more wide-band, high-resolution picture quality when possible. This value determines the memory is sufficient, the graphics resolution and refresh rate up to. If you want a resolution of 1024 × 768 to 85Hz refresh rate under, RAMDAC rate of at least 1024 × 768 × 85Hz × 1.344 (conversion factor) ≈ 90MHz. Mainstream graphics cards in 2009 to achieve 350MHz RAMDAC and 400MHz, is sufficient to meet and exceed most of the display resolution and refresh rate can provide. Need graphics power and 150 watts of electricity needed to light the same, due to operation of the IC (integrated circuits) require a lot of electricity and therefore the current generated by the internal temperature is relatively higher, so, if these temperatures were not timely reduced, then the above mentioned hardware devices are likely to suffer damage, while the cooling system is to ensure that these devices can be stable, timely operation, no heat sink or heat sink, GPU or memory will overheat, it will further damage the computer or caused when the machine, or even can not use. These cooling devices are made by the thermal conductivity of material, which some have been viewed as passive components, quiet for quiet cooling action, some are hard not to make noise such as fans. Often regarded as a passive heat sink cooling, regardless of the block is installed thermal zone, or inside other blocks, heat sinks can play its effectiveness, thus helping reduce the temperature of other devices. With heat sink and fan is usually installed to the GPU or memory, sometimes even a small fan mounted directly on the card where the highest temperatures. Graphics card is a heavy reliance on heat pipes for cooling the device, made of from ASUS Raden X 1600 to have two heat pipes, which heat can be transferred to the card slot is located behind the large heat sink for heat dissipation. The larger the surface area of ​​heat sink, has done the greater the heat dissipation efficiency (usually must work together with the fan), but sometimes because of space constraints, a large heat sink can not be installed on devices that require cooling; sometimes because the device size too small, so that bulky heat sinks can not link these devices to conduct heat. Therefore, the heat pipe to be at this time the heat from the heat at the heat transfer to the heatsink. In general, GPU heat conduction by high housing made of metal, the heat pipe will be directly linked to the chip is made of metal, this way, heat can be easily conducted to the other side of the heat sink. There are many processors, comes with heat pipe cooling system, therefore, the number of heat pipes has been developed into the cooling system can be flexibly applied in the graphics device. Most of the heat sink is just a combination of heat sink with the fan, the surface of the heatsink by the fans disperse heat, the GPU temperature is the highest part of the video cards, so cards can usually be applied to the radiator on the GPU, while the There are many accessories available to retail consumers to replace or upgrade, the most common is the VGA cooler. In order to better overclocking, overclocking is not the principle not to learn. The most effective CPU overclocking example, the current CPU production can be very precise, so that manufacturers can not control each CPU in the end what kind of frequency in the work, the manufacturers have actually done more than their tests, will be able to work at high frequencies CPU is marked as high-frequency, and can sell for a higher price. However, in order to ensure its quality, these markers have some surplus, that is, a piece of work in the 600MHZ to CPU, is likely to remain stable at 800MHZ work under, in order to explore some of these potential surplus, we can overclock. In addition, we can make use of some means to make the CPU stable operation at higher frequencies, which means mainly two things: increase the cooling effect, increase the operating voltage. Other accessories for computers, still overclock using this principle, such as the graphics card, memory, and even the mouse and so on. Do not worry, do some preparation before overclocking, these preparations will enable you to overclock could be conducted smoothly. Preparation may quicken the work, did not hurt a bit more prepared. CPU cooling fan - a very critical overclocking tool, be sure to buy a good fan, is definitely worth it! Thermal Grease - CPU and fan to increase heat transfer between the heat sink, useful things, the price is cheap. Thermal silica - General and from the chip with a small heat sink paste, to cool the motherboard chipset, graphics chip cooling, to cool the memory chips used. Small heat sink - use auxiliary cooling is mainly used to cool the chip to the heat slightly. For the avid overclocking enthusiasts, the chance of any one not to be missed overclocking, graphics card is overclocked computer can be the object of a second, naturally acclaimed, overclocked graphics card chips also depends on the core technology, the more advanced more resistant to ultra. In addition to overclocking overclocking graphics core frequency, it can also overclock memory frequency, why is the market there were a lot of video memory with 5.5ns of it? Is because the smaller the memory of the reaction time, the higher the frequency can be super, 6ns Memory can be overclocked to the general 200M, 5.5ns naturally over to the next. Overclocking memory may bring a lot of heat, we can paste the sink on the memory to alleviate this problem. [2] when the graphics card in the main work done by the coordination of the four components to complete the work, the motherboard connection equipment for the transmission of data and power, the processor used to decide how to deal with each pixel on the screen, the memory used to store information about each pixel information and the temporary storage of images has been completed, the monitor connected devices for us to see the final result. Civil graphics can be traced back in the eighties of last century had. In 1981, IBM introduced the personal computer, it offers two graphics cards, one is the , MDA is used in conjunction with the monochrome display, it can display 80 lines x25 columns of text figures, CGA can be used in the RGB display, it can draw graphics Hewen digital information. at the time speaking, the main purpose of the computer is a text data processing, although the MDA resolution wide 752 high 504 points, enough to meet the requirements of the larger display, but for word processing or have the rest of the EXCELLENCE IN GEOTECHNICAL. and CGA with color and graphics capabilities to, can qualified for general need to display graphics data, but its resolution is only 640x350, the natural color can not be displayed in the same breath. In 1982,true religion shorts melbourne Accent, IBM has launched a MGA (Monochrome Graphic Adapter), also known as Hercules Card (strong, Jacques Cartier), In addition to display graphics, but also to retain the original function of MDA. then a lot of games require this card to display the animation. The market was also popular in Genoa company to do the EGA (Enhanced Graphics Adapter), which enhanced graphics card, can simulate the MDA and CGA, but also a little bit in the monochrome screen painted graphics. EGA resolution of 640x350, 16 colors can be generated graphics and text. But these cards are based on the number of ways, until the MCGA ( Multi-Color Graphics Array) appearance, only opened a graphics card with analogue prologue. MCGA is integrated PS / 2 Model 25 and 30 on the imaging system. It uses the Analog RGA video signal, resolution up to 640x480 , digital RGB and analog RGB different places like the ON-OFF switching and fine-tune the difference between switching. display with analog RGB signals, each signal will be converted to meet the voltage range of light and dark colors. MCGA only analog display can be used together, can provide up to 256 colors, while IBM is still providing an analog monochrome display, this display can show 64 kinds of brightness. VGA (Video Graphic Array) graphics array is displayed, it is in its IBM PS / 2 of the Model 50, 60 and 80 built-in video system. It's up to 720x400 color digital mode, drawing mode, you can achieve the 640x480x16 color, and 320x200x256 color, this is the video card For the first time the highest display 256 colors simultaneously. and these models have become a common standard for all subsequent cards. VGA card into the prevalence of the computer display of the glorious era of 2D graphics. at a later period, many of the VGA card design the company continues to introduce new, the pursuit of higher resolution and bit color. At the same time, IBM introduced the 8514 / A of the Monitor screen size, is mainly used to support the 1024x 768 resolution. in the 2D era of advancing age to the process of 3D There is a can not ignore the Trident 8900/9000 graphics card, it first makes an independent accessory cards appear in the computer, instead of the integrated chip. and then the launch of Trident 9685 is the first generation of 3D graphics representatives. But the real claim to open the door to 3D graphics but should be GLINT 300SX, although the 3D feature is extremely simple, but a landmark. In 1995, for the graphics card,abercrombie polo london A person of non- Li Kan, it is definitely a milestone year, 3D graphics Accelerator passing into the player perspective. At that time the game had just entered the 3D era, the emergence of a large number of 3D games, forcing the development of the real 3D graphics accelerator card. and this year also the achievements of the company, needless to say we all know that, yes, that is, 3Dfx. 1995 年, 3Dfx, or a small company, but as a veteran of the 3D technology company, he launched the industry's first piece of the true meaning of 3D graphics accelerator card: Voodoo. the most popular games at the time Kennedy in the motor, Voodoo in speed and color performance of all users who like crazy games, many game enthusiasts have had to take more than a thousand dollars in it to the computer to buy a piece of the city's Voodoo graphics-name experience. 3Dfx Glide patented technology engine interface, once the kings of the whole 3D world, until the D3D and OpenGL, before we changed this situation. Voodoo comes standard with 4Mb memory, can provide 640 × 480 resolution in the 3D display speed and the most gorgeous screen, of course, Voodoo is also flawed, it is just a child with a 3D acceleration card is required to use features with a 2D graphics card has, I believe many of the old players are eligible EDO S3 765 + Voodoo remember the man Jin-Jin Le Road gold portfolio. talked about S3 765, S3 would have mentioned that the former king of graphics card. S3 765 video card is compatible then the standard configuration, up to 2MB EDO video memory, to achieve high-resolution display, which at the time is the efficacy of high-end graphics card, SVGA to really flourish in this chip. can support 1024 × 768 resolution, and low resolution support up to 32Bit true color, and the price is also strong. Therefore, S3 765 S3 Graphics in effect brought the first glory. and then in 96 years and launched the S3 Virge, it is a combination of 3D graphics acceleration, support for DirectX, and contains many of the advanced 3D acceleration, such as Z-buffering, Doubling buffering, Shading , Atmospheric effect, Lighting, actually become a pioneer in 3D graphics, the achievements of the S3 graphics card's second brilliant, but unfortunately under the catch-up later in the 3Dfx, S3 Virge series no longer following the brilliant, the market eventually abandoned. Since then, in order to repair Voodoo is not 2D display this flawed, 3Dfx and then launched a VoodooRush, which joined the Z-Buffer technology, unfortunately opposed to Voodoo, VoodooRush the 3D performance without any increase, even more frightening is bring a lot of compatibility problems and the high price of factors restricted the promotion of VoodooRush card. Of course, when the 3D graphics card market is not 3Dfx hoodwink the public, high prices to other companies left many living space, such as survey said cost was the King Trident 9750/9850, and provides hardware decoding Mpeg-II technology, SIS6326, also the first time in the history of cards played in the nVidia launched Riva128/128zx, have been a lot of favorite players, which also contributed to the graphics card technology development and market maturity. burst on the scene in 1997 was the first 3D graphics feet a year, and 1998 is a 3D graphics card a year of intense competition springing up. 2098, 3D gaming to the Bay from the wind, a lot more beautiful mass market 3D games, allowing users and vendors are looking forward to faster and more powerful graphics card there. In Voodoo tremendous honor and numerous accolades, 3Dfx to a commanding height of the potential launch of yet another landmark product: Voodoo2. Voodoo2 from with 8Mb/12Mb EDO memory, PCI interface, the card has dual-chip, multi-texture can be done single-cycle operation. Voodoo2 course, there are shortcomings, it is the card itself is very long, and the chip heat is very large, has also become a worry, and Voodoo2 3D acceleration is still as a daughter card, need a 2D graphics support. But it is undeniable, Voodoo2 has led to the introduction of 3D acceleration has reached a new milestone, with the Voodoo2 effects, graphics and speed, conquered a lot of time temporary popular 3D games, such as Fifa98, NBA98,the true religion womens, Quake2 and so on. Maybe a lot of users do not know, the most popular in 2009 when Voodoo2 SLI technology is a new technology, Voodoo2 support dual graphics technology for the first time, so two Voodoo2 work in parallel to double the performance. Voodoo2 1998, although the shine of the year, but other manufacturers are also some classics. Matrox MGA G200 in the succession of his own other than the standard super-class 2D, 3D aspects of a revolutionary improve, not only can provide and Voodoo2 similar processing speed and special effects, in addition to support for DVD hardware decoding and video output, and pioneered a unique 128-bit dual independent bus, greatly improving the performance, was very popular with the AGP bus , G200 is also the favorite to win a lot of users. Intel's I740 is paired with Intel's 440BX chipset was launched, it supports AGP 2X technology, standard 8Mb memory, I740 Unfortunately, the performance is not good, 2D performance can only and S3 Virge line, and also only Riva128 3D levels, but prices have obvious advantages, it stood at the low end feet. Riva TNT is nVidia launched products aimed at blocking Voodoo2, which comes standard with 16Mb of large memory, full support for AGP technology, the first support 32-bit color rendering, as well as faster than the D3D Voodoo2 Voodoo2 performance and lower the price, make it a new favorite of many players. and have been battles in the Apple world, ATI has produced a product called Rage Pro graphics card, slightly faster speed than Voodoo. IDF Spring 2001, Intel officially announced on PCI Express, PCI bus to replace the third-generation I / O technology, also known as 3GIO. The bus specification supported by the Intel AWG (Arapahoe Working Group) is responsible for. April 17, 2002, AWG announced 3GIO 1.0 specification finalized the draft, and the transfer of PCI-SIG for review. the beginning we all thought it would be named Serial PCI (subject to the impact of the serial ATA), but in the end but was officially named as PCI Express. 2006, officially launched Spec2.0 (2.0 specification). PCI Express bus technology evolution, in fact, computing system I / O interface rate the process of evolution. PCI bus is a 33MHz @ 32bit or 66MHz @ 64bit parallel bus, the bus bandwidth of 133MB / s to a maximum of 533MB / s, PCI bus connected to all devices share the 133MB / s ~ 533MB / s bandwidth. This bus is used to cope with the sound card, 10/100M LAN, and USB 1.1 interface, such as basic is not a problem. With the computer and communications technology, further development, a new generation of I / O interfaces in large numbers, such as Gigabit (GE), Gigabit (10GE) Ethernet technology, 4G/8G of FC technology, making the PCI bus bandwidth, computing systems have been unable to cope with concurrent read and write within a large number of high-bandwidth requirements, PCI bus has become a bottleneck in system performance, so there have been PCI Express bus. PCI Express bus technology in next-generation storage system has universal application. PCI Express bus to provide high bandwidth to meet the system requirements. As of 2009, PCI-E 3.0 specification has also been identified, the coding data rate, under the same circumstances than the PCI-E 2.0 specification doubled, X32 port bi-directional speeds up to 320Gbps. PCI bus, the bus structure, the biggest advantage is that simple, low cost, simple design, but the disadvantages are obvious: 1) parallel bus can not connect too many devices, bus expansion is rather poor, the interference will cause the system to line does not work; 2) When connecting multiple devices, the bus effective bandwidth will be greatly reduced, the transmission rate slow; 3) To reduce costs and to minimize interference between each other, need to reduce the bus bandwidth, or the address bus and data bus by way of design reuse, which reduces the bandwidth utilization. PCI Express bus is the future of computers and the definition of a high-performance communications platform , General I / O interconnect bus. Compared with the PCI bus, PCI Express bus technology are the following advantages: 1) is a serial bus, point to point transmission,fivefingers shoes Search Engine Optimisation SEO -, exclusive bandwidth for each transmission channel. 2) PCI Express bus supports bi-directional transmission mode and data sub-channel transfer mode. in which the data transfer mode that sub-channel PCI Express bus x1, x2, x4, x8, x12, x16 and x32 multi-channel connectivity, x1 can be achieved by one-way transmission bandwidth 250MB / s, two-way transmission bandwidth is more able to achieve 500MB / s, this is not an ordinary PCI bus, which can be compared to the. 3) PCI Express bus full use of advanced point to point interconnect, based on the exchange of technology, packet-based protocol implement new bus performance and features. power management, quality of service (QoS), Hot Plug support, data integrity, error handling mechanism is also supported by the PCI Express bus, advanced features. 4) and PCI bus, a good inheritance, inheritance and to maintain the reliability of the software. PCI Express bus, the PCI key features, such as the application model, storage structure, software interfaces, consistent with the traditional PCI bus, but the parallel PCI bus is a highly scalable, fully serial replaced by bus lines. 5) PCI Express bus full use of advanced point to point interconnections, reducing system design complexity of hardware platforms and difficulty, thus greatly reducing the cost of manufacturing design system development, greatly increase system cost-effective and robustness. It can be seen from the table below, the system bus bandwidth while reducing the number of hardware PIN, directly down the cost of the hardware. to 2008, PCI-E graphics interface is still in use.