home / Installation and configuration/ Features of national modernization, or what is not worth saving on. Video card, how to choose

Features of national modernization, or what is not worth saving on. Video card, how to choose

Modern graphics processors contain many functional blocks, the number and characteristics of which determine the final rendering speed, which affects the comfort of the game. By the comparative number of these blocks in different video chips, you can roughly estimate how fast one or another GPU is. Video chips have a lot of characteristics, in this section we will consider only the most important of them.

Video chip clock frequency

The operating frequency of a GPU is usually measured in megahertz, that is, in millions of clock cycles per second. This characteristic directly affects the performance of the video chip - the higher it is, the more work the GPU can perform per unit of time, process large quantity vertices and pixels. An example from real life: the frequency of the video chip installed on the Radeon HD 6670 is 840 MHz, and the exact same chip in the Radeon HD 6570 runs at 650 MHz. Accordingly, all the main performance characteristics will differ. But not only the operating frequency of the chip determines the performance, its speed is strongly influenced by the graphics architecture itself: the device and the number of execution units, their characteristics, etc.

In some cases, the clock speed of individual GPU blocks differs from the clock speed of the rest of the chip. That is, different parts of the GPU operate at different frequencies, and this is done to increase efficiency, because some units are capable of operating at higher frequencies, while others are not. Most of these GPUs come with GeForce video cards from NVIDIA. From recent examples, we present a video chip in the GTX 580 model, most of which operates at a frequency of 772 MHz, and the universal computing units of the chip have a doubled frequency - 1544 MHz.

Filling rate (fill rate)

The fill rate shows how fast the video chip is capable of rendering pixels. There are two types of fillrate: pixel fill rate and texel rate. Pixel fill rate shows the speed at which pixels are drawn on the screen and depends on the operating frequency and the number of ROPs (blocks of rasterization and blending operations), and texture is the speed of fetching texture data, which depends on the frequency of operation and the number of texture units.

For example, the peak pixel fill rate of the GeForce GTX 560 Ti is 822 (chip clock) × 32 (ROP units) = 26304 megapixels per second, and texture fill rate is 822 × 64 (number of texture units) = 52608 megatexels / s. In a simplified way, the situation is as follows - the larger the first number, the faster the video card can render the finished pixels, and the larger the second, the faster the texture data is sampled.

Although the importance of "pure" fill rate has dropped noticeably lately, giving way to the speed of computations, these parameters are still very important, especially for games with simple geometry and relatively simple pixel and vertex computations. So both parameters remain important for modern games, but they need to be balanced. Therefore, the number of ROP units in modern video chips is usually less than the number of texture units.

The number of computational (shader) units or processors

Perhaps, now these blocks are the main parts of the video chip. They perform special programs known as shaders. Moreover, if earlier pixel shaders executed blocks of pixel shaders, and vertex ones - vertex blocks, then for some time graphic architectures were unified, and these universal computing units began to deal with various calculations: vertex, pixel, geometric and even universal calculations.

For the first time, a unified architecture was applied in a video chip game console Microsoft Xbox 360, this GPU was developed by ATI (later purchased by AMD). And in video chips for personal computers unified shader units appeared in the NVIDIA GeForce 8800 board. And since then, all new video chips are based on a unified architecture, which has a universal code for different shader programs (vertex, pixel, geometric, etc.), and the corresponding unified processors can execute any program ...

By the number of computing units and their frequency, one can compare the mathematical performance of different video cards. Most games are now limited by the performance of pixel shaders, so the number of these units is very important. For example, if one model of a video card is based on a GPU with 384 computational processors in its composition, and another from the same line has a GPU with 192 computational units, then with an equal frequency the second will process any type of shader twice as slow, and in general it will be the same more productive.

Although it is impossible to draw unambiguous conclusions about performance solely on the basis of the number of computing units alone, it is imperative to take into account the clock frequency and different architecture of units of different generations and chip manufacturers. Only these figures can be used to compare chips only within the same line of one manufacturer: AMD or NVIDIA. In other cases, you need to pay attention to performance tests in the games or applications of interest.

Texture mapping units (TMU)

These GPU units work in conjunction with computational processors to select and filter texture and other data required for scene construction and general-purpose computing. The number of texture units in the video chip determines the texture performance - that is, the speed of fetching texels from textures.

Although recently more emphasis has been placed on mathematical calculations, and some of the textures are being replaced by procedural ones, the load on TMUs is still quite high, since in addition to the main textures, selections must also be made from normal and displacement maps, as well as off-screen render target buffers.

Taking into account the emphasis of many games, including on the performance of texturing units, we can say that the number of TMUs and the corresponding high texture performance are also one of the most important parameters for video chips. This parameter has a special effect on the rendering speed of the image when using anisotropic filtering, which require additional texture selections, as well as with complex soft shadow algorithms and newfangled algorithms like Screen Space Ambient Occlusion.

Rasterization Operations Blocks (ROPs)

Rasterization units perform operations of recording pixels calculated by the video card into buffers and operations of their mixing (blending). As we noted above, the performance of ROP units affects the fill rate and this is one of the main characteristics of video cards of all time. And although recently its value has also slightly decreased, there are still cases when application performance depends on the speed and the number of ROPs. This is most often due to the active use of post-processing filters and the enabled anti-aliasing at high game settings.

Once again, we note that modern video chips cannot be evaluated only by the number of different blocks and their frequency. Each GPU series uses a new architecture, in which the execution units are very different from the old ones, and the ratio of the number of different units may differ. Thus, AMD ROP units in some solutions can perform more work per clock than NVIDIA units, and vice versa. The same applies to the capabilities of TMU texture units - they are different in different GPU generations. different manufacturers, and this must be taken into account when comparing.

Geometric Blocks

Until recently, the number of geometry processing units was not particularly important. One GPU block was enough for most tasks, since geometry in games was quite simple and the main focus of performance was mathematical calculations. The importance of parallel processing of geometry and the number of corresponding blocks has grown dramatically with the advent of support for geometry tessellation in DirectX 11. NVIDIA company was the first to parallelize the processing of geometric data, when several corresponding blocks appeared in its chips of the GF1xx family. Then, a similar solution was released by AMD (only in the top solutions of the Radeon HD 6700 line based on Cayman chips).

Within the framework of this material, we will not go into details, they can be read in the basic materials of our site dedicated to DirectX 11-compatible graphics processors. What is important for us in this case is that the number of geometry processing units greatly affects the overall performance in the newest games that use tessellation, such as Metro 2033, HAWX 2 and Crysis 2 (with the latest patches). And when choosing a modern gaming video card, it is very important to pay attention to geometric performance.

Video memory size

Video chips use their own memory to store the necessary data: textures, vertices, buffer data, etc. It would seem that the more there is, the better. But not everything is so simple, estimating the power of a video card by the amount of video memory is the most common mistake! Inexperienced users often overestimate the value of video memory, still using it for comparison. different models video cards. It is understandable - this parameter is indicated in the lists of characteristics of ready-made systems one of the first, and on the boxes of video cards it is written in large print. Therefore, it seems to an inexperienced buyer that since the memory is twice as large, then the speed of such a solution should be twice as high. Reality differs from this myth in that memory is different types and characteristics, and the growth of productivity grows only up to a certain amount, and after reaching it, it simply stops.

So, in every game and with certain settings and game scenes there is a certain amount of video memory, which is enough for all data. And even though you put 4 GB of video memory there, it will have no reason to accelerate rendering, the speed will be limited by the execution units, which were discussed above, and there will simply be enough memory. That is why, in many cases, a video card with 1.5 GB of video memory operates at the same speed as a card with 3 GB (all other things being equal).

There are situations where more memory leads to a visible increase in performance - these are very demanding games, especially at ultra-high resolutions and at maximum settings quality. But such cases do not always occur and the amount of memory must be taken into account, not forgetting that the performance simply will not increase above a certain amount. Have memory chips and more important parameters, such as the width of the memory bus and its operating frequency. This topic is so extensive that we will dwell on the choice of video memory in more detail in the sixth part of our material.

Memory bus width

Memory bus width is the most important characteristic affecting throughput memory (PSP). The wider width allows more information to be transferred from the video memory to the GPU and back per unit of time, which has a positive effect on performance in most cases. Theoretically, a 256-bit bus can transfer twice as much data per clock as a 128-bit bus. In practice, the difference in rendering speed, although it does not reach two times, is very close to this in many cases with an emphasis on video memory bandwidth.

Modern gaming graphics cards use different bus widths: from 64 to 384 bits (previously there were chips with a 512-bit bus), depending on the price range and release time specific model GPU. For the cheapest video cards of the low-end level, 64 and less often 128 bits are most often used, for the middle level from 128 to 256 bits, but video cards from the upper price range use buses from 256 to 384 bits wide. The bus width can no longer grow purely due to physical constraints - the size of the GPU die is not enough to lay out more than a 512-bit bus, and this is too expensive. Therefore, the memory bandwidth is now being increased by using new types of memory (see below).

Video memory frequency

Another parameter that affects memory bandwidth is its clock frequency. And increasing memory bandwidth often directly affects the performance of a video card in 3D applications. The memory bus frequency on modern video cards ranges from 533 (1066, doubling) MHz to 1375 (5500, quadrupling) MHz, that is, it can differ by more than five times! And since the memory bandwidth depends both on the memory frequency and on the width of its bus, memory with a 256-bit bus operating at 800 (3200) MHz will have a higher bandwidth compared to memory operating at 1000 (4000) MHz with a 128-bit bus.

Particular attention should be paid to the parameters of the memory bus width, its type and operating frequency when buying relatively inexpensive video cards, many of which are equipped with only 128-bit or even 64-bit interfaces, which has an extremely negative effect on their performance. In general, we do not recommend buying a video card using a 64-bit video memory bus for a gaming PC at all. It is advisable to give preference to at least an average level with a minimum of 128- or 192-bit bus.

Memory types

Several different types of memory are installed on modern video cards at once. Old SDR memory with a single baud rate is nowhere to be found anywhere, but also modern types DDR memory and GDDR have significantly different characteristics. Various types DDR and GDDR allow you to transfer two or four times more data at the same clock frequency per unit of time, and therefore the figure of the operating frequency is often indicated double or quadrupled, multiplied by 2 or 4. For example, if the frequency of 1400 MHz is indicated for DDR memory , then this memory operates at a physical frequency of 700 MHz, but indicate the so-called "effective" frequency, that is, the one at which the SDR memory must operate in order to provide the same bandwidth. The same is with GDDR5, but the frequency is even quadrupled here.

The main advantage of the new types of memory is the ability to work at high clock speeds, and, accordingly, in an increase in bandwidth compared to previous technologies. This is achieved due to increased latencies, which, however, are not so important for video cards. The first board to use DDR2 memory was NVIDIA GeForce FX 5800 Ultra. Since then, graphics memory technologies have advanced significantly, the GDDR3 standard has been developed, which is close to the DDR2 specifications, with some changes specifically for video cards.

GDDR3 is a memory specially designed for video cards, with the same technologies as DDR2, but with improved consumption and heat dissipation characteristics, which made it possible to create microcircuits operating at higher clock frequencies. Despite the fact that the standard was developed by ATI, the second modification of the NVIDIA GeForce FX 5700 Ultra was the first video card to use it, and the next was the GeForce 6800 Ultra.

GDDR4 is further development"Graphics" memory, which runs almost twice as fast as GDDR3. The main differences between GDDR4 and GDDR3, significant for users, are once again increased operating frequencies and reduced power consumption. Technically, GDDR4 memory doesn't differ much from GDDR3 memory, it is a further development of the same ideas. The first video cards with GDDR4 chips on board were ATI Radeon X1950 XTX, while NVIDIA did not release products based on this type of memory at all. The advantages of the new memory chips over GDDR3 are that the power consumption of the modules can be about a third lower. This is achieved at the cost of a lower voltage rating for GDDR4.

However, GDDR4 is not widely used even in AMD solutions. Starting with the RV7x0 GPU family, video card memory controllers support a new type of GDDR5 memory, operating at an effective quadruple frequency up to 5.5 GHz and higher (theoretically possible frequencies up to 7 GHz), which gives a bandwidth of up to 176 GB / s using 256-bit interface. While the GDDR3 / GDDR4 memory had to use a 512-bit bus to increase the memory bandwidth, the switch to GDDR5 made it possible to double the performance with smaller die sizes and lower power consumption.

Video memory of the most modern types- these are GDDR3 and GDDR5, it differs from DDR in some details and also works with double / quadruple data transfer. These types of memory use some special technologies to increase the operating frequency. So, GDDR2 memory usually operates at higher frequencies than DDR, GDDR3 - at even higher, and GDDR5 provides the maximum frequency and bandwidth for this moment... But inexpensive models are still equipped with “non-graphic” DDR3 memory with a much lower frequency, so you need to choose a video card carefully.

GPU architecture: features

The realism of 3D graphics is highly dependent on the performance of the video card. The more blocks of pixel shaders the processor contains and the higher the frequency, the more effects can be applied to a 3D scene to improve its visual perception.

The GPU contains many different functional blocks. By the number of some components, you can estimate how powerful the GPU is. Before moving on, let me review the most important functional blocks.

Vertex processors (vertex shader units)

Like pixel shader units, vertex processors execute shader code that touches vertices. Since a larger vertex budget allows you to create more complex 3D objects, the performance of vertex processors is very important in 3D scenes with complex objects or a large number of them. However, vertex shader units are still not as obviously affecting performance as pixel processors.

Pixel Processors (Pixel Shader Units)

A pixel processor is a component of the graphics chip dedicated to processing pixel shader programs. These processors perform pixel-only computations. Because pixels contain color information, pixel shaders can achieve impressive graphical effects. For example, most of the water effects you've seen in games are created using pixel shaders. Typically, the number of pixel processors is used to compare the pixel performance of video cards. If one card is equipped with eight pixel shader units and the other with 16 units, then it is quite logical to assume that a video card with 16 units will process complex pixel programs faster. You should also consider the clock speed, but today doubling the number of pixel processors is more energy efficient than doubling the frequency of the graphics chip.

Unified shaders

Unified shaders have not yet arrived in the PC world, but the upcoming DirectX 10 standard is based on a similar architecture. That is, the structure of the code of vertex, geometric and pixel programs will be the same, although the shaders will perform different work. The new spec can be viewed on the Xbox 360, where the GPU was specially designed by ATi for Microsoft. It will be very interesting to see what potential new DirectX 10.

Texture Mapping Units (TMU)

Textures should be selected and filtered. This work is done by the texture mapping units, which work in conjunction with the pixel and vertex shader units. The TMU's job is to apply texture operations to the pixels. The number of texture units in GPU often used to compare the texture performance of video cards. It is quite reasonable to assume that a video card with a large number TMU will give higher texture performance.

Raster Operator Units (ROPs)

RIPs are responsible for writing pixel data into memory. The rate at which this operation is performed is the fill rate. In the early days of 3D accelerators, ROPs and fill rates were very important characteristics of graphics cards. Today, ROP performance is still important, but the performance of a video card is no longer limited by these blocks, as it used to be. Therefore, the performance (and number) of ROPs is already rarely used to estimate the speed of a video card.

Conveyors

Pipelines are used to describe the architecture of video cards and provide a very visual representation of the performance of the GPU.

Conveyor is not a strict technical term. The GPU uses different pipelines that perform different functions. Historically, a pipeline was understood as a pixel processor that was connected to its own texture mapping unit (TMU). For example, in Radeon video cards The 9700 uses eight pixel processors, each of which is connected to its own TMU, so the card is considered to have eight pipelines.

But it is very difficult to describe modern processors by the number of pipelines. Compared to previous designs, the new processors use a modular, fragmented structure. ATi can be considered an innovator in this area, which switched to modular structure, which made it possible to achieve performance gains through internal optimization. Some CPU blocks are used more than others, and to improve the performance of the GPU, ATi tried to find a compromise between the number of blocks required and the die area (it cannot be increased too much). In this architecture, the term "pixel pipeline" has lost its meaning, since pixel processors are no longer connected to their own TMUs. For example, the GPU ATi Radeon The X1600 has 12 pixel shader units and a total of four TMUs. Therefore, one cannot say that the architecture of this processor has 12 pixel pipelines, just like say that there are only four of them. However, by tradition, pixel pipelines are still mentioned.

Taking these assumptions into account, the number of pixel pipelines in a GPU is often used to compare video cards (with the exception of the ATi X1x00 line). For example, if we take video cards with 24 and 16 pipelines, then it is quite reasonable to assume that a card with 24 pipelines will be faster.

CONTENT

What will be discussed in this short article?

This article is a set of basic knowledge for those who want to choose a balanced graphics card without giving extra money to marketers. Will help beginners, as well as serve as a source useful information and for more advanced PC users. Nevertheless, the mini article is nevertheless focused on for newbies.

The purpose of the video card.

It's no secret that in our time, the main field of activity for productive graphics card are - 3 Dgames, smooth playing video( HD ), work in professional 3D2D and video editors. Rest, daily tasks can be performed without any problems on video cards built into the processor or chipset. Recently, the field of activity for the video card has been expanded, in the form multithreaded computing that run much faster on a parallel graphics card architecture than on processors.

NVidiapromotes its software and hardware platformCUDAbased on language Si (by the way, it is successful, and this is not surprising, when investing such funds).AMDhowever, mostly relies on open sourceOpenCL.

By using can encode video in 3-4 times faster... Hardware acceleration of the company's products by means of video cardsAdobe- in particular Photoshop, Flashand this is apparently just the beginning. True, those people who constantly use the computing power of video cards are theoretically very few. And it seemed too early to think about it, especially since they are stepping on the heels manynuclear processors, which, although slower in multi-threaded operations, have an indisputable advantage in that they simply do their job without complex software optimizations. And the simplicity and ease of implementation, as history showsWindows(for example) - for people the main thing and the key to success at Software market. Still, it is worth paying tribute to the computing power of video cards, which is not yet tamed by the "correct" software.

So. NVidiaorAMD?

* The most "interesting" question

The main players in the graphics accelerator market are corporationsAMD and NVidia.

Everything is clear here, as in many market sectors, duopoly. How Pepsi and Coca - Cola like Xbox 360 , how Intel and AMD eventually. Recently, companies have been releasing their products one by one. So that both one was good and the second. At first AMD releases the flagship of the line, then after two or three months, more powerful flagship releases NVidia... First, cards are bought from AMD like the most powerful, then after the cards are released NVidia who bought them go to the store again for an even better product. Almost the same thing happens with the medium and budget markets. Only the spread in the increased performance relative to the competitor is higher here, since in order to interest a more economical consumer, something more is required than a chance to have a better video card, as is the case in the flagship sector.

It is better not to be fanatical, because this is business and nothing personal. The main thing is that the video cards are productive, and the prices do not bite. And which manufacturer is not important. With this approach, you can always win in terms of price performance.

Chip architecture.

Quantitypixel processors (for AMD ), universal conveyors (for NVidia).

Yes. These are completely different things. What AMD has Radeon HD 5870 – 1600 executive blocks does not mean at all that it will be 3 times more powerful thanNVidia GTX 480 which has on board 480 executive blocks.

NVidiaIt has scalar architecture, andAMD– super scalar .

AMD architecture.

Consider the architecture PP (* pixel processors),on the example of the basic super scalar architecture of video cardsRadeon HD 5 series ( 5-way VLIW).

Every 5 pp make up one executive block, which at a time can execute at most - 1 scalar operation and 1 vector or sometimes 5 scalar(however, the conditions are not always suitable for this). Each vector operation requires 4 PP, each scalar 1 PP... And then, how it goes. HaveNVidia same, each Cuda core, performs strictly according to 1 vector and 1 scalar operations per clock cycle.

With the release of the 6th episode, under the codename ( Nothern islands ), namely Cayman chips, decided to abandon the additional, fifthALU(T-unit), who was responsible for performing complex tasks.

Now this role can be played by three of the four remaining blocks. This allowed offloading the thread manager ( Ultra-Threaded Dispatch Processor), which also doubled to improve geometry and tessellation, which were the weaknesses of the 5 series. Plus, it allows you to save on core area and transistor budget with the same efficiency.

After the sixth series, work towards development VLIW ended due to its weak flexibility and long downtime due to the dependencies of internal blocks from each other (in particular vector operations). A completely new architecture has come to the fore Graphics Core Next .

Engine SIMD, is replaced by a computing unit Compute Unit (CU), which can significantly raise the level of efficiency and performance of the architecture. Each PP can now independently perform vector and scalar operations, since separate control blocks have been introduced for them, which more efficiently distribute resources between free blocks. In general, architecture begins to acquire some prerequisites for scalar architecture from NVidia which is simple and efficient.

The first chip with the new architecture was GPU Tahiti on which are built AMD Radeon HD 7970/7950 ... The company plans to release and middle class on the new architecture.

Now let's look at the basic one, scalar architecture NVidia .

As we can see, each universal processor ( ), per beat performs 1 scalar operation and 1 vector. This allows for maximum smoothness. Where there are many vector and scalar operations, video cardsAMD with architecture VLIWinferior, since they are not able to load their blocks with work like video cardsNVidia.

Let's say the choice fell betweenRadeon HD 5870 and GeForce GTX 480 .

The first 1600pp, the second 480 unified blocks.

Calculate: 16005 = 320 superscalar blocks, y Radeon HD 5870.

That is, for a clock cycle, a video card fromAMD, performs from 320 to 1600 scalar operations and from 0 to 320 floating vector, depending on the nature of the problem.

And at the doubled frequency of the shader domain, the card on the architectureFermi, theoretically should perform 960 vector and 960 scalar operations per clock cycle.

but Radeon , has a better frequency than the green camp card (700 versus 850). So, such indicatorsNVidia, theoretically, they should be the same as when the shader domain operates at 1700 MHz (850 x 2 = 1700), but it is not. At a frequency of 1401 MHz, GTX 480 produces ~ 700 vector and ~ 700 scalar operations per clock cycle.

* Do not rely on the accuracy of these calculations, they are only theoretical. In addition, this statement does not apply from the 6th series. Radeon starting with chips Cayman.

Due to the fact that the maximum number of vector and scalar operations is the same number, the architectureNVidiahas the best smoothness in complex scenes than AMD VLIW (<5 series).

Price categories and what we get if we buy a video card in a series of younger ones.

Engineers AMDwithout hesitation, they cut half of the pixel processors, the memory bus and partROP’S generation of cards, from segment to class below. For exampleRadeon HD5870 It has 1600pp, tire 256 bitand in 577 0, exactly half of this remains - 800 , and the memory bus 128 bit... The same situation continues up to the most budget video cards. So, it is always preferable to purchase a weaker graphics card from the 58 ** series than the older one from the 57 ** series.

Engineers NVidia, not much different approach. Smoothly, memory bus is clipped, universal pipelines,ROP’S , pixel pipelines. But the frequencies also decrease, which, with a proper cooling system, can be slightly compensated for by overclocking. It's a little strange that it's not the other way around, as it doesAMD, increasing the frequencies on cards with a cut off number of executive elements.

An approach AMD more profitable for the manufacturer, the approach NVidia- to the buyer.

Mention of drivers.

It is because of the peculiarities of the superscalar architecture VLIW, drivers from AMD, you have to constantly optimize so that the video card understands when it needs to use vectors or scalars as efficiently as possible.

Unified drivers fromNVidiamore immune to different game engines, due to the fact that engineersNVidiaoften already during the development of a game, they optimize it for the architecture of their video chips and drivers. It is also worth noting that when installing and removing them, there are practically no problems that are inherent in drivers fromAMD.

Drivers NVidia can be installed directly on old ones, without uninstalling and without cleaning the registry. Hope programmersAMDwill move in the same direction. Now you can download "fixes" for driversCatalyst, which are released shortly before the release of the game or a little later. Already something. And with the release of a new architecture Graphics Core Next, the driver optimization work will be much easier.

Pixel conveyors, TMU, ROP.

Also, the number is very important pixel pipelines and TMU (texture mapping unit), their number is especially important at high resolutions and when using anisotropic texture filtering ( pixel pipelines are important), using high quality textures and high anisotropic filtering settings (important TMU).

Number of blocksROP (raster operations blocks ), mainly affect the performance of anti-aliasing, but if they are lacking, there may be a loss in overall performance. The more there are, the more imperceptible anti-aliasing will affect the number of frames per second. Also, the performance of anti-aliasing is significantly affected by the amount of video memory.

Volume, frequency and width of the memory bus.

The more video memory a video card has, the better. However, it is not worth buy a lot.

As often happens, on relatively weak video cards, they put incredible amounts of video memory, and even slow (for example,GeForce 8500 GT, some OEMmanufacturers put on 2 GB DDR2 video memory). From this, the video card will not take off, and performance will not be added.

* compared to 8500 GT 512 MB

A much better option would be to take a video card with faster memory, but less volume. For example, if the choice is worth: take 9800 GTwith 512 or 1024 mb memory, with a frequency 1000mhz and 900MHz accordingly, it would be preferable to take 9800 GT with 512 mb memory. Moreover, a video card of this level does not need more video memory than 512 mb.

Memory bandwidth - this is the main thing in the performance of the video memory subsystem, which in the most important way affects the performance of the video card as a whole. Measured in GB / s (gigabytes per second).

For example now, video memory likeGddr5 , which has a much higher frequency potential thanGddr3 , and, accordingly, whiter than high bandwidth.

However, frequency is not everything. The second important factor is memory bus width. The higher the bit depth, the faster the memory.

For example, memory with frequency 1000mhz and bus 256 bit, will be exactly 2 times faster memory 1000mhz and bus 128 bit... The more bit depth, the faster the memory. The widest memory bus in existence is a monstrous 896 bit(448 x2 ) on the video card GeForce GTX295 ... However, it uses memoryGddr3 , which significantly degrades the bandwidth (less effective frequency) in comparison withGddr5 ... Therefore, its bandwidth is even slightly lower than that ofRadeon HD 5970 with 512 bit(256 x 2), but with Gddr5 .

Cooling system.

The more efficient the cooling system, the less chance your graphics card will fail. The card will overheat less, which will improve the overall system stability, significantly increase life time, as well as increase overclocking potential.

Manufactured, ready-madewithsystems O There are two variations of cooling for video cards.

Reference (from the manufacturer) and alternative (from the manufacturer's partners). As a rule, reference cards are of a turbine (, blower) design, and are usually very reliable. Relatively noisy, not always as effective as alternative CO from the manufacturer's partners and become more clogged with dust. Although in use, the video card bleeding systems are very efficient and quiet. If a little noise during load does not bother you, and you will not set records in overclocking, reference cooling systems are preferable. Usually, partners of manufacturers paste over them with stickers with their logos, changes are possible only in the BIOS of the video card (fan speed control), therefore some cards are identical in design, but from different manufacturers, noisier or hotter than their counterparts and vice versa. Each manufacturer has its own preferences and warranty conditions. Therefore, some sacrifice silence for greater stability and durability.

If it is important to you silence, then you should pay attention to alternative systems cooling with increased efficiency, with a lower noise level (for exampleVapor - x, IceQ, , DirectCu), or choose a video card with a passive cooling system, of which there are more and more now.

* Advice: do not forget to change the thermal interface once a year or two, especially for CO with the technology of direct contact of heat pipes. Thermal grease hardens, forming a layer that does not conduct heat well, which leads to overheating of the video card.

Power consumption of the video card.

A very important characteristic when choosing, since a video card is a very gluttonous component of a computer, if not the most gluttonous. Top graphics cards sometimes approach the mark 300W... Therefore, when choosing, you should consider whether your power supply is capable of providing a stable power supply to the video card. Otherwise, the system may either not start due to a voltage mismatch when passing POST, instability and unexpected shutdowns, reboots or overheating of computer components may appear, or the power supply may simply burn out.

On the manufacturer's website or on the box of the video card, the minimum specifications are written, including the minimum power of the power supply. These values are written for any blocks, including Chinese ones. If you are sure that you have a high-quality power supply, you can subtract from this value 50-100W.

You can indirectly determine the power consumption by the number of additional power connectors on the video card.

None less 75W, one 6-pin before 150W, two 6-pin before 225W, 8-pin + 6-pin - before 300W... Make sure that your unit has the necessary connectors or that the kit includes adapters for 4-pin molex-NS. Or buy them in addition, they are freely sold in computer stores.

Lack of power supply to the video card can lead to its overheating, artifacts and the failure of its power system. Video cards NVidia, if there is a lack of power, they may start warning messages of the form: "the video driver stopped responding and was restored" or "connect additional power to the video card".

High power consumption = high heat dissipation... If your graphics card consumes a lot of power, take care of additional blowing and blowing fans on the case. Or, as a temporary measure, open the side cover. Constantly high temperature in the case has a detrimental effect on the service lines of all components from the motherboard to the end.

Connectors.

When you have already decided on a video card, you should also pay attention to the connectors.

If you have a monitor with a matrix P- or with support 30 bit color (1.07 billion), then you will definitely need DisplayPort on a video card to unleash its potential. Only DisplayPort supports transmission 30 bit depth of color.

* It is not known for certain whether gaming video cards support 30-bit transmission, but the presence DisplayPort talks about possible support. In the specifications, support is declared only for professional video cards AMD FirePro and NVidia Quadro.

It is very good if there is ... You never know what might come in handy and it's better to be ready for it. Suddenly you need to output the signal from the receiver. By the way, HDMI and DVI compatible via a simple adapter and virtually no problem.

Conclusions.

That's all. We did not have time to start, we are already finishing. Since the article describes the main, general concepts, it turned out to be not too long.

Nevertheless, all the most important points for choosing a high-quality and productive video card are described.

1. A question of faith.

3. The number of execution units (TMU, ROP, etc.).

4. Volume, frequency and capacity of the memory bus.

5. Find out if the card is suitable for the level of power consumption.

5. Cooling system.

6. Connectors.

We hope that with this knowledge, you will be able to choose a video card according to your requirements.

Good luck with your choice!

Perhaps, now these blocks are the main parts of the video chip. They run special programs known as shaders. Moreover, if earlier pixel shaders executed blocks of pixel shaders, and vertex ones - vertex blocks, then for some time graphic architectures were unified, and these universal computing units began to deal with various calculations: vertex, pixel, geometric and even universal calculations.

The unified architecture was first used in the video chip of the Microsoft Xbox 360 game console, this GPU was developed by ATI (later acquired by AMD). And in video chips for personal computers, unified shader units appeared in the NVIDIA GeForce 8800 board. And since then, all new video chips are based on a unified architecture, which has a universal code for different shader programs (vertex, pixel, geometric, etc.), and the corresponding unified processors can execute any program.

Texture mapping units (TMU)

Rasterization Operations Blocks (ROPs)

On our forum, dozens of people every day ask for advice on the modernization of their own, in which we willingly help them. Every day, "evaluating the assembly" and checking the components selected by our customers for compatibility, we began to notice that users are mainly paying attention to other, no doubt, important components. And rarely does anyone remember that when upgrading a computer, it is imperative to update an equally important detail -. And today we will tell and show you why you shouldn't forget about it.

“… I want to upgrade my computer, everything was flying, I bought a percentage of i7-3970X and an ASRock X79 Extreme6 motherboard, plus a RADEON HD 7990 6GB vidyahu. What else nan ???? 777 "
- this is how about half of all messages related to updating a stationary computer begin. Based on their own or family budget, users try to choose the most, most and the fastest and most beautiful memory modules. At the same time, naively believing that their old 450W will cope with both a gluttonous video card and a "hot" processor during overclocking at the same time.

For our part, we have already written about the importance of the power supply unit more than once - but, we confess, it was probably not clear enough. Therefore, today we have corrected ourselves, and have prepared for you a memo on what will happen if you forget about when upgrading your PC - with pictures and detailed descriptions.

So, we decided to update the configuration ...

For our experiment, we decided to take a completely new average computer and upgrade it to the level of a "gaming machine". You won't have to change the configuration too much - it will be enough to change the memory and the video card so that we have the opportunity to play more or less modern games with decent detail settings. The initial configuration of our computer is as follows:

Power Supply: ATX 12V 400W

It is clear that this configuration is rather weak for games, to put it mildly. So it's time to change something! We'll start with the same thing that most people eager for an "upgrade" start with - p. We will not change the motherboard - as long as it suits us.

Since we decided not to touch the motherboard, we will select an FM2 compatible socket (fortunately, for this there is a special button on the NIKS website on the motherboard description page). Let's not be greedy - let's take an affordable but fast and powerful processor with a frequency of 4.1 GHz (up to 4.4 GHz in Turbo CORE mode) and an unlocked multiplier - we also love to "overclock", nothing human is alien to us. Here are the specs for our chosen processor:

Specifications

CPU bus frequency

5000 MHz

Power dissipation

100 watts

Processor frequency

4.1 GHz or up to 4.4 GHz in Turbo CORE mode

Core

Richland

L1 cache

96 KB x2

L2 cache

2048 KB x2, clocked at processor frequency

64 bit support

Yes

Number of Cores

Multiplication

41, unlocked multiplier

Processor video core

AMD Radeon HD 8670D @ 844 MHz; Shader Model 5 support

Max RAM

64 GB

Max. number of connected monitors

3 direct-connected or up to 4 monitors using DisplayPort splitters

One 4GB bar is not our choice. Firstly, we want 16GB, and secondly, we need to use a two-channel mode of operation, for which we will install two memory modules with a capacity of 8GB each in our computer. High throughput, lack of radiators and a decent price make these the most "tasty" choice for us. In addition, from the AMD website, you can download the Radeon RAMDisk program, which will allow us to create a super-fast virtual drive up to 6GB for free absolutely free - and everyone loves free useful things.

Specifications
Memory	8 GB
Number of modules	2
Memory standard	PC3-10600 (DDR3 1333 MHz)
Functioning frequency	up to 1333 MHz
Timings	9-9-9-24
Supply voltage	1.5V
Bandwidth	10667 Mb / s

You can comfortably play the embedded video only as a "sapper". Therefore, in order to upgrade the computer to a gaming level, we chose a modern and powerful, but not the most expensive one.

She became with 2GB of video memory, support for DirectX 11 and OpenGL 4.x. and the excellent Twin Frozr IV cooling system. Its performance should be more than enough for us to enjoy the latest parts of the most popular gaming franchises like Tomb Raider, Crysis, Hitman and Far Cry. The characteristics of the chosen one are as follows:

Specifications
GPU	GeForce GTX 770
GPU frequency	1098 MHz or up to 1150 MHz in GPU Boost mode
Number of shader processors	1536
Video memory	2 GB
Video memory type	GDDR5
Video memory bus width	256 bit
Video memory frequency	1753 MHz (7.010 GHz QDR)
Number of pixel pipelines	128, 32 texture sampling units
Interface	PCI Express 3.0 16x (PCI Express 2.x / 1.x compatible) with SLI card interconnection.
Ports	DisplayPort, DVI-D, DVI-I, HDMI, D-Sub adapter included
Cooling the graphics card	Active (radiator + 2 Twin Frozr IV fans on the front side of the board)
Power connector	8 pin + 8 pin
API support	DirectX 11 and OpenGL 4.x
Graphics card length (measured in NIKS)	263 mm
Supports general-purpose GPU computing	DirectCompute 11, NVIDIA PhysX, CUDA, CUDA C ++, OpenCL 1.0
Maximum power consumption FurMark + WinRar	255 watts
Performance rating	61.5

Unexpected difficulties

Now we have everything we need to upgrade our computer. We will install new components in our existing case.

We launch it - and it does not work. And why? But because budget power supplies are not physically capable of starting a computer with any slightest degree. The fact is that in our case two 8-pin connectors are required for power supply, and the power supply unit has only one 6-pin video card power connector "in the base". Considering that many more need even more connectors than in our case, it becomes clear that the power supply needs to be changed.

But this is not so bad. Just think, there is no power connector! In our test lab, we found quite rare adapters from 6-pin to 8-pin and from molex to 6-pin. Like these ones:

It is worth noting that even on budget modern power supplies, Molex connectors are getting smaller with each new release - so we can say we were lucky.

At first glance, everything is fine, and with some tweaks we were able to update the system unit to a "gaming" configuration. Now let's simulate the load by running Furmark and 7Zip in Xtreme Burning on our new gaming machine at the same time. We could start the computer - that's good. The system also survived the Furmark launch. We start the archiver - and what is it ?! The computer turned off, having delighted us with the roar of the fan turned to maximum. The "modest" standard 400W failed, no matter how hard it tried, to feed the video card and powerful processor. And because of the mediocre cooling system, ours got very hot, and even the maximum fan speed did not allow it to deliver at least the declared 400W.

There is an exit!

They sailed. We bought expensive components to assemble a gaming computer, but it turns out that it’s impossible to play on it. It's a shame. The conclusion is clear to everyone: the old one is not suitable for our gaming computer, and it urgently needs to be replaced with a new one. But which one?

For our upgraded computer, we chose according to four main criteria:

The first is, of course, power. We preferred to choose with a margin - we would also like to overclock the processor and gain points in synthetic tests. Taking into account everything that we may need in the future, we decided to choose a power of at least 800W.

The second criterion is reliability... We really want the one taken "with a margin" to survive the next generation of video cards and processors, not burn itself out and at the same time not burn expensive components (along with the test site). Therefore, our choice is only Japanese capacitors, only short-circuit protection and reliable overload protection for any of the outputs.

The third point of our requirements is convenience and functionality.... To begin with, we need - the computer will work often, and especially noisy PSUs, coupled with a video card and a processor cooler, will drive any user crazy. In addition, the sense of beauty is no stranger to us, so a new power supply for our gaming computer should be modular and have detachable cables and connectors. So that there is nothing superfluous.

And last but not least, the criterion is energy efficiency... Yes, we care about the environment and our electricity bills. Therefore, the power supply we choose must meet at least the 80+ Bronze energy efficiency standard.

After comparing and analyzing all the requirements, we chose among the few applicants that fully satisfied all our requirements. It became a power of 850W. Note that in a number of parameters it even surpassed our requirements. Let's see its specification:

Power supply specifications
Type of equipment	Power supply with active PFC (Power Factor Correction) module.
Properties	Loop braiding, Japanese capacitors, Short circuit protection (SCP), Overvoltage protection (OVP), Overload protection of any of the unit outputs separately (OCP)
+ 3.3V - 24A, + 5V - 24A, + 12V - 70A, + 5VSB - 3.0A, -12V - 0.5 A
Detachable power cables	Yes
Efficiency	90%, 80 PLUS Gold Certified
Power supply unit	850 watts
Motherboard power connector	24 + 8 + 8 pin, 24 + 8 + 4 pin, 24 + 8 pin, 24 + 4 pin, 20 + 4 pin (collapsible 24-pin connector. 4-pin can be detached if necessary, collapsible 8-pin connector)
Video card power connector	6x 6/8-pin connectors (collapsible 8-pin connector - 2 pins detachable)
MTBF	100 thousand hours
Cooling the power supply	1 fan: 140 x 140 mm (on the bottom wall). Passive cooling system up to 50% load.
Fan speed control	From a thermal sensor. Changing the fan speed depending on the temperature inside the power supply. Manual selection of the fan operating mode. In Normal mode, the fan rotates continuously, and in Silent mode, it stops completely at low load.

, one of the best for the money. Let's install it in our corpus:

Then something happened that confused us a little. It would seem that everything was assembled correctly, everything was connected, everything worked - but the power supply is silent! That is, in general: the fan is still standing still, and the system is properly started and functioning. The fact is that at a load of up to 50%, the power supply operates in the so-called quiet mode - without spinning the cooling fan. The fan will hum only under heavy load - the simultaneous launch of archivers and Furmark made the cooler spin.

The power supply has as many as six 8-pin6-pin video card power connectors, each of which is a collapsible 8-pin connector, from which, if necessary, you can unfasten 2 contacts. Thus, it is able to feed any video card without unnecessary hassle and difficulties. And not even one.

The modular power supply system allows you to detach unnecessary and unnecessary power cables, which improves the airflow of the case, the stability of the system and, of course, aesthetically improves the appearance of the internal space, which allows us to safely recommend to modders and fans of cases with windows.
buy a reliable and powerful power supply. In our review, it became. - and as you can see, it is no coincidence. By purchasing one from NICS, you can be sure that all the components of your high-performance system will be provided with sufficient and uninterrupted power, even with extreme overclocking.

In addition, the power supply will last for several years in advance - it is better with a margin, in case you are going to update the system with high-level components in the future.