Device description amd radeon hd 7800 series. AMD Radeon video card families Reference information. Overclocking and temperatures

7800 Series with features that differ from each other. The chip, created on the Graphic Core Next microarchitecture, occupies a space equal to 2.8 billion transistors. Like most Radeon cards, it features Eyefinity technology, which allows you to connect up to six monitors simultaneously. They can work independently of each other, or they can form one large monitor. It all depends on what settings will be set.

Radeon 7850

This AMD 7800 Series video card has 800 megahertz processor frequency. High performance and throughput (153 gigabits per second) are provided by a 256-bit bus. The computing system processes data equal to 1.76 teraflops. There are 16 computational units, and 64 texture units. There are two cores for computing processes.

The memory format corresponds to the GDDR5 marking, and support for DirectX version 11 will help speed up interaction with operating system applications. To better optimize the performance of the card, it is necessary to monitor driver updates, since only they can fully unlock all the capabilities of the GPU and provide access to the necessary settings. Basic drivers that identify the video card in the system are included with the card, and the updated version can be viewed on the AMD website.

This AMD Radeon HD 7800 Series graphics processor supports the latest built-in technologies that allow you to enjoy high-quality and smooth images in 60 frames, and the resolution can reach 4096 x 2160 pixels. The same applies to the audio stream, which meets all modern requirements, producing high-quality sound.

Radeon 7870

This AMD Radeon HD 7800 Series video card is a powerful successor to the previous card. It has a whole gigahertz at its disposal to work with the graphics processor. The performance for computing operations is much higher than in the previous version - 2.56 teraflops. There are 20 computational units and 80 texture units.

Since this is the flagship of the 7800 series, it is superior to its brother in many respects. Support for tessellation technology has been implemented in video cards from this manufacturer for a long time, but in this version it is taken to the limit. Now you can enjoy a three-dimensional image that is amazing in its realism and detail. And improved anti-aliasing will help achieve a smooth and pleasant picture.

In other parameters, this representative of the AMD Radeon HD 7800 Series is completely identical in characteristics to the previous video card. Both cards are capable of supporting 3D technology in both video and games. It is also possible to connect multiple cards to improve the performance, but this parameter may also depend on the capabilities of the motherboard.

Chip codename: "Tahiti"
4.3 billion transistors (more than 60% more than the Cayman and exactly twice as many as the Cypress)
384-bit memory bus: six 64-bit wide controllers supporting GDDR5 memory
Core clock: up to 925 MHz (for Radeon HD 7970)
32 GCN compute units, including 128 SIMD cores, consisting of a total of 2048 floating point ALUs (integer and floating point formats, support for FP32 and FP64 precision within the IEEE 754 standard)
128 texture units, with support for trilinear and anisotropic filtering for all texture formats
32 ROP units with support for anti-aliasing modes with programmable sampling of more than 16 samples per pixel, including with FP16 or FP32 frame buffer format. Peak performance up to 32 samples per clock, and in Z only mode - 128 samples per clock
Integrated support for six monitors, including HDMI 1.4a and DisplayPort 1.2

Radeon HD 7970 graphics card specifications

Core frequency: 925 MHz
Number of universal processors: 2048
Number of texture blocks: 128, blending blocks: 32
Effective memory frequency: 5500 MHz (4x1375 MHz)
Memory type: GDDR5
Memory capacity: 3 gigabytes
Memory bandwidth: 264 gigabytes per second.
Theoretical maximum fill rate: 29.6 gigapixels per second.
Theoretical texture sampling rate: 118.4 gigatexels per second.
Two CrossFire connectors
PCI Express 3.0 bus
Power consumption: from 3 to 250 W
One 8-pin and one 6-pin power connector
Dual slot design
Recommended price for the US market: $549

Radeon HD 7950 graphics card specifications

Core frequency: 800 MHz
Number of universal processors: 1792
Number of texture blocks: 112, blending blocks: 32
Effective memory frequency: 5000 MHz (4x1250 MHz)
Memory type: GDDR5
Memory capacity: 3 gigabytes
Memory bandwidth: 240 gigabytes per second.
Theoretical maximum fill rate: 25.6 gigapixels per second.
Theoretical texture sampling rate: 89.6 gigatexels per second.
Two CrossFire connectors
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, two Mini-DisplayPort 1.2
Power consumption: from 3 to 200 W
Dual slot design
US MSRP: $449

The high complexity of the new chip is noteworthy - 4.3 billion transistors, which is more than half the number of transistors in the previous top-end graphics processor. The ability to make such a complex crystal was made possible by the use of a modern 28-nanometer process technology, and the new chip was even slightly smaller in area than the Cayman. And its practical characteristics that affect performance have been noticeably improved: the number of ALUs, TMUs, and memory bus. Only the number of ROP blocks did not increase, and the frequency of GDDR5 video memory remained at the same level.

The naming principle for the company's video cards remains the same. The Radeon HD 7970 is the company's most productive single-chip solution; after some time, the junior HD 7950 model was released, announced a little later. Initially, the HD 7970 had no competitors on the market and did not replace any specific video card from the AMD line, but rather moved it down. As for comparison with its competitor, NVIDIA released its 28-nanometer solution much later.

The new AMD video card is equipped with the same GDDR5 memory, but its volume, instead of 2 gigabytes in the previous generation, has increased to 3 gigabytes. This happened due to the expansion of the memory bus from 256-bit to 384-bit. And now you can put either 1.5 GB or 3 GB on the new board. Naturally, from a marketing point of view, installing a smaller volume would be a clear disadvantage, so the decision was made to install 3 GB, although today this is a bit overkill. Only in ultra-high resolutions and with MSAA 16x is 1.5-2 GB not enough. However, AMD also has Eyefinity, and for games on three or more monitors, the screen buffer will take up a very large volume.

So, let's look at the Radeon HD 7970. The new video card in the upper price range has a dual-slot cooling system, covered with a plastic casing familiar to all modern AMD boards along the entire length of the card. Only the design of this casing has changed a little, although the back part still extends beyond the printed circuit board. But the design of the strip with pins was changed - to improve cooling of the video card, one of the two slots (half of the strip) was occupied exclusively by a ventilation hole for heat removal.

But users should not suffer from a reduction in the number of DVI connectors soldered directly on the board. For their convenience, a special HDMI-DVI adapter will be included in the package, which will allow you to connect two monitors with DVI connectors. By the way, the power consumption of the new card is no lower than that of the Radeon HD 6970, so it had to be equipped with a set of one 8-pin and one 6-pin power connector.

But in the new Radeon HD 7970 the cooling system has changed for the better. A new generation of evaporation chamber and a new larger cooler are used, with reshaped blades and increased performance (more air flow is provided). The result is an increase in cooler efficiency while reducing noise.

The Dual BIOS firmware switch, which we wrote about in the description of the Radeon HD 6900, has also not disappeared from the board. In short: the video card has two BIOS versions, one with the ability to custom flash the firmware, and the second with firmware hardcoded at the factory. Both users and AMD itself liked this convenient solution so much that it decided to continue to package them with top-end solutions.

We can only welcome this solution, which really helps in various cases related to unexpected problems during flashing (power outage during the process, for example), and allows you to fearlessly carry out various experiments with BIOS images. It is not surprising that AMD again and again hints at excellent overclocking capabilities of the new video card:

As you can see, overclocking to a frequency of 1 GHz and higher is practically promised, if you do not take into account the small inscription (which was not included in the screenshot) that the warranty ceases to be valid even if the video card fails as a result of an experiment with raising the frequency from the settings video drivers.

Architectural features of the Radeon HD 7970

To appreciate the relevance of the architectural modifications in Southern Islands, we first look at the development of GPUs over the past few years as represented by AMD. Until 2002, graphics chips were specific hardware capable of exclusively graphics computing. Video chips of that time had limited functionality; they could only apply and filter textures, process geometry, and perform primitive rasterization and therefore were not at all suitable for universal computing tasks.

Over the next few years, basic programmability was added to the GPU, but also focused exclusively on graphics tasks. This was the time of support for DirectX 8 and 9, limited-function shader programs with the ability to calculate with floating point. Video chips of that time had specialized ALU units for vertex and pixel processing, as well as dedicated caches for pixels, textures and other data. Versatility was still not even close.

It was only in 2007 that AMD acquired the unified DirectX 10 shader architecture, as well as the ability to program the GPU using special tools: CAL, Brook, ATI Stream. GPUs of that time already had advanced caching and support for local and global shared data. Architecturally, the chips were based on VLIW5 and VLIW4 blocks, flexible enough for some basic non-graphics calculations, but still focused on graphics algorithms.

Now it's time for a new architecture that's even better suited for general purpose computing - Graphics Core Next (GCN). This is a new architectural era for AMD, which is why the name was chosen. The new GPUs offer excellent graphics processing capabilities and performance, but the architectural changes made are intended primarily to improve positions in non-graphics computing - increasing performance and efficiency in complex general-purpose tasks. The new GPU design is intended for so-called heterogeneous computing - a mixture of graphics and general purpose computing in a multitasking environment. The GCN architecture has become more flexible and should be even better suited for energy-efficient execution of various tasks.

The basic block in the new architecture is the GCN block. It is on these “building blocks” that all new Southern Islands series graphics processors are based. The architecture for the first time for AMD graphics chips uses a non-VLIW design, it uses vector and scalar units, and one of the most important changes is that each of the GCN computing units has its own scheduler and can execute instructions from various programs (kernel).

The new computing architecture is designed for high efficiency in loading compute units in a multi-tasking environment. The GCN computational unit is divided into four subsections, each of which works on its own instruction stream every clock cycle. Threads can also use the scalar block provided by GCN for flow control or pointer operations. The combination of vector and scalar blocks offers a very simple programming model. For example, function pointers and stack pointers are much easier to program, and the compiler's task is now significantly simplified, since the execution units are scalar.

Each GCN block has 64 KB of dedicated local data storage for data exchange or local register stack expansion. The block also includes a first-level cache with read and write capabilities, and a full-fledged texture pipeline (sampling and filtering units). Therefore, the new computing unit is able to work independently, without a central scheduler, which in previous architectures was responsible for distributing work across units. Now each of the GCN blocks is capable of scheduling and distributing commands itself; one computing block can execute up to 32 different command streams, which can be from different virtual address spaces in memory and are completely protected and independent of each other.

Previous AMD GPU architectures used the VLIW4 and VLIW5 architectural models, and although they are good enough for graphics tasks, they are not efficient enough for general-purpose computing, since it is very difficult to load all the execution units with work in such conditions. The new GCN architecture offers the same large number of execution units, but with a scalar execution that removes the limitations and dependencies of registers and instructions. The transition from the VLIW architecture to scalar execution provides a noticeable simplification of code optimization tasks.

When executing instructions on the previous VLIW4 architecture, the compiler has to deal with register conflicts, perform complex distribution of instructions to execution units at the code compilation stage, etc. At the same time, achieving high performance often requires non-trivial optimization, which is suitable for most graphics tasks and much less flexible for other calculations. The new architecture offers significant simplification of development and support, simplified creation, analysis and catching of errors in low-level code, stable and predictable performance.

Memory caching subsystem

There is never enough bandwidth and memory and caches, and there is always a need and methods to increase them. AMD's new GPUs use a full two-level read/write cache. Each computing unit has 16 kilobytes of first-level cache, and the total volume of second-level cache is 768 kilobytes (in total, the chip has 512 KB L1 and 768 KB L2), which is 50% more than in the previous chip, which does not have write capabilities at all in L2 cache.

In terms of performance, each GCN compute unit can receive or write 64 bytes of data from/to the L1 cache or global memory, which is used to exchange data between instruction threads, in one clock cycle. Each section of the second level L2 cache is capable of transmitting and receiving the same amount of data. As a result, the company's top-end GPU achieves 2 terabytes/s for L1 and 700 GB/s for L2, which is 50% more than AMD's previous top-end solution.

Tahiti GPU

Now that we've looked at the low-level architectural changes in the new Southern Islands series, it's time to move on to the details of the line's most powerful solution, the Radeon HD 7900, which includes two models. First of all, let's note the sheer complexity of the new GPU, because it includes more than 4.3 billion transistors, which is twice as many as were in the chip on which the Radeon HD 5870 is based! Naturally, such a powerful chip became possible only thanks to the use of the new 28 nm process technology. So what does he have inside?

The number of geometric blocks has not changed, compared to the Cayman, there are still two of them, but the efficiency of their work has been significantly increased - we will dwell on this in more detail a little later. In the GPU diagram we see 32 GCN architecture compute units available on the Radeon HD 7970, and in the case of the lower-end solution, some of them will be disabled. If we consider the peak computing performance of the solution, it is almost 3.8 teraflops (floating point operations per second), which is an absolute record for a GPU to date.

Each GCN block contains 16 texture units, which gives a final figure of 128 TMUs per chip, or more than 118 gigatexels/sec - and this is another record at the time of release, and it will not be the last. But the number of ROP blocks has not changed, there are still 32 of them in 8 enlarged RBE blocks. Another interesting architectural change is that ROP blocks are now “attached” not to memory channels, as was previously the case, but to GCN blocks.

Although theoretically the write speed to the framebuffer has remained almost unchanged, and the maximum possible is the same 32 color values and 128 depth values per clock, the practical fill rate in real-world applications has increased significantly due to the increased memory bandwidth. According to AMD's measurements, the Cayman recorded only 23 pixels per clock, while the new Tahiti came close to the theoretical 32 pixels per clock.

This is understandable, because AMD's new video chip has a 384-bit memory bus - six 64-bit channels, just like the competitor's current top-end solution. It is this one and a half times increase in memory bandwidth that makes it possible to increase the actual speed of texture fetching and writing to the framebuffer. Bandwidth of 264 GB/sec should help to squeeze out close to theoretical figures of 118 gigapixels/sec and 30 gigapixels/sec, and in the practical part we will check this.

In the case of the “cut down” Radeon HD 7950 GPU, Tahiti includes 28 active computing units of the GCN architecture out of 32 physically available on the chip. In the case of the junior solution of the Radeon HD 7970 series, it was decided to disable four of them. Since each GCN unit contains 16 texture units, the total number of TMUs for the new model is 112 TMUs, which gives a performance of almost 90 gigatexels/sec.

But the number of ROP units and memory controllers in the HD 7950 has not changed; they decided not to cut them and leave them at the same 32 and 6 pieces, respectively. Therefore, the Tahiti Pro video chip has the same 384-bit memory bus, assembled from six 64-bit channels, as the top AMD solution. Apparently, it is the computing functional devices that suffer the most from production defects, and they decided not to cut everything else.

Tessellation and geometry processing

From an architectural point of view, nothing much has changed about the Tahiti's geometric blocks since the Cayman. Two blocks are still used for processing (setting vertices and tessellation) geometric data and rasterization, and the scheme is very similar to what we saw earlier, except that the tessellators are called the 9th generation:

Despite the schematic similarities, the latest generation of these blocks are capable of significantly greater tessellation and geometry processing performance, as the blocks have undergone significant modifications. Although peak performance only increased to almost two billion vertices and primitives per second (925 MHz and two vertices per clock), actual performance increased more. This was achieved by increasing the size of caches, improving geometry data buffering, and reusing vertex data.

As a result, tessellation performance is improved across all triangle ratios by up to four times compared to the previous generation Radeon HD 6970. But four times are not achieved in all cases, even on the diagram from AMD itself:

The chart compares the tessellation performance of the Radeon HD 7970 versus the HD 6970 at partitioning factors ranging from 1 to 32. And as you can see, the performance difference is between 1.7 and 4 times. But this is bare synthetic. And to get closer to reality, let’s provide some more data on tessellation speed in gaming applications:

As you can see, AMD's synthetic numbers are well supported by gaming ones - performance in real applications with “heavy” tessellation has increased significantly. This is a very good result, which we will definitely check in the practical part, using the example of synthetics and gaming applications.

Non-graphical computing

From the point of view of heterogeneous and non-graphical computing tasks, the emergence of two asynchronous computing engines (Asynchronous Compute Engines - ACE) is very important. They are designed to schedule and distribute work between execution units for efficient multitasking and work in conjunction with the graphical command processor (Command Processor).

Radeon HD 7900 has two independent computing engines and one graphics engine. In total, this gives three programmable blocks and three command streams, completely separate from each other. And in addition to asynchronous command delivery for fast context switching, the new GPU also features two bidirectional direct memory access (DMA) controllers introduced in the Cayman. These two controllers are required to take full advantage of the new PCI Express 3.0 bus.

As we know, from the point of view of serious computing, not only the speed of performing single precision floating point operations is important, but also double precision floating point. And the new AMD architecture copes with this task quite well. At the moment, it is assumed that there are two versions of GCN computational units that have different rates of execution of FP64 instructions. For older GPUs, the execution rate is 1/4 of the FP32 speed, and for younger chips the execution rate is 1/16, which is quite enough to maintain compatibility, but does not complicate inexpensive solutions too much. As a result, the Radeon HD 7970 is capable of 947 billion double-precision operations per second (oh, they just fell short of a teraflop!) - this is another highest achievement of the new AMD chip.

Moreover, these are not the same gigaflops as in the case of previous architectures, but more “fat” ones. After all, the efficiency of the new GPU in complex computing tasks should increase significantly. Firstly, the memory and caching subsystem has been improved. Second, each GCN compute unit has its own scheduler, which should improve branching code execution and overall efficiency. Well, thirdly, we note the scalar execution, which does not require complex optimizations from the compiler, as a result of which the computational units will be idle much less often. And as a result, in any tasks it will be easier for the new chip to demonstrate high performance and ALU load.

Among other innovations related to computing capabilities, we note full ECC support for DRAM and SRAM. On the software side, it is important that Tahiti is the first GPU with full support for the new API versions: OpenCL 1.2, DirectCompute 11.1 and C++ AMP and their capabilities. For example, OpenCL 1.2 allows you to combine the capabilities of multiple computing devices into one, and AMD has already released support for this in the form of AMD APP SDK 2.6 and the Catalyst 11.12 driver.

Architecture performance and efficiency

After reviewing all the architectural innovations using the example of the top-end chip of the Southern Island series, it’s time to talk about the effectiveness of all these changes. It is clear that the performance of the new chips is much higher than that of the previous ones; the opposite would be quite surprising. The question is how much faster. In various tasks, the results range from 40-50% (minimum!) to a fivefold difference. Improvements in the architecture allow it to exceed the theoretical 1.4 times the difference in stupid gigaflops. Let's look at this with examples:

The diagram compares the new top solution and the previous single-chip solution: Radeon HD 7970 and HD 6970, which is quite fair. The selected performance tests are different: SmallptGPU and LuxMark are ray tracing on OpenCL, SHA256 is a secure hashing algorithm, and AES256 is a symmetric encryption algorithm. Well, Mandelbrot is a well-known problem calculated with double precision.

The vertical dashed line in the graph marks the theoretical difference in performance, but the speed data shows that in three out of five tasks the speed of the new GPU was significantly higher. This is caused by all the changes aimed at increasing efficiency: moving away from VLIW, having a scheduler in each computing unit, improved caching, etc.

Changes to rendering quality

Actually, this part could easily have been skipped, since lately there have been no special complaints about the image quality and there cannot be any - for various reasons. For example, the quality of full-screen anti-aliasing among video cards from different manufacturers is very similar, especially considering the widespread use of software anti-aliasing methods using post-processing filters, which are performed exactly the same on all GPUs.

The same applies to texture filtering - now its quality is such that it is very difficult to distinguish between AMD and NVIDIA solutions even if you make a pixel-by-pixel comparison. In the Radeon HD 6900 - the company's previous generation - anisotropic filtering has improved a little more, and now even a “microscope” will not help to find any significant shortcomings there. The only note is that in motion the Radeon video cards were slightly inferior to the GeForce due to more noticeable specific artifacts, such as “noise” or “sand”.

With the release of new generation video chips, the texel weights in the texture filter were reviewed again, modifying them in such a way as to reduce such artifacts, sometimes visible on the Radeon HD 6900 in the presence of certain types of textures (“high-frequency”, with sharp transitions from dark to light, for example). The changes in quality are so difficult to show with examples that AMD does not provide comparative pictures of the HD 7900 versus the HD 6900, but simply compares the quality of the “hardware” algorithm with a purely software one executed on GPU stream processors, and therefore ideal:

In such a small screenshot, the difference in quality is not visible, but AMD assures that all the changes made did not introduce any drop in performance and did not degrade the picture quality in any aspect - it still does not depend on the angle and the filtering quality is close to ideal. We will definitely check this in one of the future practical materials.

Partially Resident Textures

The idea of Partially Resident Textures (PRT) is to use the hardware capabilities of the presented GPU - virtual memory. Surely many users have already seen the game RAGE by id Software, which uses virtual texturing technology, the so-called megatexturing (“MegaTexture”), which makes it possible to use huge amounts of texture data and stream them into video memory.

Using virtual video memory, it is very easy to get effective hardware support for such algorithms, allowing you to use up to 32 terabytes of textures in an application, which makes it possible to create unique locations in games, without repeating pieces of textures, and with no problems with loading texture data. True, AMD gives a clear example that is too strange, from which nothing is particularly clear:

PRT allows you to achieve high picture quality and helps improve the efficiency of video memory use. Similar algorithms are already used in the id Software engine, and they are expected to appear in many next-generation engines. Games of the future need to work with huge amounts of data, and the advantage of the new GPU is that local graphics memory in algorithms a la PRT works as hardware cache memory, and textures are loaded into it when necessary. GPUs of the Southern Islands family support “megatextures” with a volume of up to 32 terabytes (resolution up to 16384 × 16384) and, most importantly, hardware texture filtering for them, which is not available on earlier video chips.

Virtual textures are divided into pieces of 64 kilobytes in size (kilobytes, not texels) and this piece size is fixed. And only those that are needed when rendering the current frame are loaded into the local memory of the video card. The technology works regardless of the texture format, just the sizes of the pieces in texels will be different. For example, for a regular uncompressed texture with 32 bits per color, the piece size will be 128x128 texels, and for a texture compressed in DXT3 format - 256x256 texels.

The technology also involves the use of texture mip levels (reduced copies used in texture filtering). They require multiple accesses during rendering and filtering. Let's look at the operation of the algorithm using an example.

This image highlights four different chunks from different mip levels required for rendering. When a shader program requests data from them, some of the chunks are already in local memory and this data is immediately sent to the shader for further calculations. But some chunks are missing from the table, and the application must choose what to do next on that miss. For example, you can request data from a lower resolution mip level, then the image will be fuzzy, but at least it will look like the real thing and will be rendered without delay. And by the time the next frame is rendered, it can already be loaded into the cache - local video memory. Those who played RAGE will understand us.

This is a powerful algorithm that allows you to use huge textures that are unique to each object. Similar algorithms have long been used in offline rendering, except for the need for real-time calculations. AMD even made a demo using the Per-Face Texture Mapping technique developed by Walt Disney Animation Studios for their animated films. Unfortunately, the demo isn't ready yet, and we've only seen low-resolution screenshots.

The essence of this texture mapping technique is to assign a specific piece of texture to each polygon, without the need to use UV transformation (finding a correspondence between the surface coordinates of a three-dimensional object and the coordinates on a two-dimensional texture). This approach solves some of the problems with creating tessellated content by making the displacement mapping algorithm very simple. And PRT in this method is used to efficiently store and access texture data.

Media Processing Instructions

An interesting innovation in Southern Islands seems to be support for specialized instructions used in image processing, static and dynamic. For example, a widely used instruction called “sum of absolute differences,” better known as SAD (Sum of Absolute Differences), has been improved. The speed of its execution is a very critical performance bottleneck for many image and video processing algorithms, such as motion detection, gesture recognition, image search, computer vision and many others.

But in our review of the ancient Radeon HD 5870 video card, we already wrote about SAD support. Now, in addition to the regular SAD (4x1), Southern Islands has a new instruction - QSAD (quad SAD), which combines SAD with shift operators to increase performance and energy efficiency, as well as a “masked” instruction MQSAD, which ignores background pixels and is used for isolation objects moving in the frame from the background.

The new GPUs can process up to 256 pixels per GCN compute unit per clock, which in the case of the AMD Radeon HD 7970 model means the ability to process up to 7.6 trillion pixels per second in the case of 8-bit integer color values. Although this is a theoretical figure, the visual processing capabilities of the new GPUs are quite impressive - many video processing tasks can be performed in real time.

PCI Express 3.0

We couldn't ignore the support of the third version of PCI Express by the entire line of new Southern Islands graphics solutions. This support was quite expected, since the specifications of the third version of PCI Express were finally approved back in the fall of 2010, but there were still no hardware solutions with its support, although motherboards are already appearing, video cards were released at the end of 2011, and the corresponding central processors There is.

The updated interface has a transfer speed of 8 gigatransactions per second instead of 5 GT/s for version 2.0, and its throughput has once again doubled (to 32 GB/s) compared to the PCI Express 2.0 standard. The new bus uses a different encoding scheme for data sent over the bus, but compatibility with previous versions of PCI Express has been maintained.

The first motherboards supporting PCI Express 3.0 were introduced in the summer of 2011, mainly based on the Intel Z68 chipset, and they became widely available only in the fall of the same year. Now the video cards have arrived, and AMD has once again become ahead of the rest in terms of the speed of release of new graphics processors supporting the most advanced technologies. But whether PCI-E 3.0 will make any practical sense is too early to judge.

AMD PowerTune Technology

One of the most interesting innovations in the Cayman was PowerTune advanced power management technology. Flexible GPU power management has been used for a long time, but before the Radeon HD 6900, all these technologies were rather primitive and mostly software methods and changed the frequency and voltage in steps, unable to turn off large parts of video chips.

Even in the Radeon HD 5000 family, a performance limiter appeared when a certain consumption level was exceeded, and in the Radeon HD 6900 the system moved to a qualitatively different level. To do this, the chip included special sensors in all blocks that monitor boot parameters. The GPU constantly measures load and power consumption and does not allow the latter to go beyond a certain threshold, automatically adjusting the frequency and voltage so that the parameters remain within the specified thermal package.

Unlike earlier power management technologies, PowerTune provides direct control over GPU power consumption, as opposed to indirect control by changing frequencies and voltages. This technology helps you set your GPU clocks to high speeds, achieving high gaming performance without worrying about power consumption going beyond safe limits. After all, most games and general applications that use GPU computing have significantly lower power requirements and do not approach dangerous power consumption limits, unlike stability tests like Furmark and OCCT.

Even the heaviest games do not require maximum power consumption, and if you limit consumption by frequency, testing video cards with extreme tests, then in the case of 3D games there will be quite a lot of untapped performance and power potential. In the case when the video card has not reached the safe consumption level, the GPU will operate at the frequency set at the factory, and in the FurMark and OCCT tests, the GPU frequency will decrease to remain within the consumption limits.

Thus, PowerTune helps set higher factory frequencies and configure the system to make the most efficient use of GPU resources at the set maximum consumption level. In the example shown above, the HD 5870 does not use PowerTune and, due to the GPU frequency limitation due to high consumption in the endurance tests, does not use all its capabilities. While the Radeon HD 7970 is set to the maximum TDP, and the video chip resets frequencies only when it is exceeded, obtaining the highest possible performance in any application.

This is clearly shown in the following diagram. In the case of gaming applications, achieving TDP is possible by increasing the GPU frequency, and for peak loads, endurance tests reduce the frequency to a safe level of power consumption. Without PowerTune, you would have to choose - either the possibility of a video card failure when using FurMark and OCCT for a long time, or reduce the potential performance in games. New technology solves these issues as efficiently as possible.

AMD PowerTune is fast responsive to changing conditions (microseconds) because it is a hardware technology. It is also distinguished by flexible frequency settings, and not stepwise, as was the case in previous chips. All measurements are independent of the driver, but can be adjusted by the user using the video card settings.

The difference between PowerTune and the previously generally accepted approach is that in other cases, thermal throttling is used, which puts the GPU into a significantly reduced consumption mode, while PowerTune simply smoothly reduces its frequency, bringing GPU consumption to the set limiter. This achieves higher clock speeds and performance.

AMD ZeroCore Technology

AMD did not limit itself to using power management technology already known from previous solutions. In the first chips in the Southern Islands family, it introduces AMD ZeroCore technology, which helps achieve even greater energy efficiency in the "deep idle" (or "sleep") mode with the display device disabled, which is supported by all operating systems.

After all, almost any system, even a gaming one, spends most of its time in low load mode on the graphics processor. And the video card should not consume a lot of energy in this mode. And even more so, not to mention the mode with the monitor turned off - in this case, it is advisable to turn off the GPU completely. That's what AMD did. Thanks to ZeroCore, in a deep idle state, the new GPU consumes less than 5% of the energy of a full mode, disabling most functional blocks in this mode.

AMD provides a schematic comparison with its own Radeon HD 5870, which did not support such technology. ZeroCore is a Southern Islands exclusive innovation brought to desktop solutions from mobile GPUs designed for laptops. By the way, the advantages of this technology are associated not only with reduced consumption. In addition, in long-term idle mode when the display is turned off, the video card also completely turns off the fan on the video card cooler!

This is exactly what many users have been waiting for for a long time. The most interesting thing is that, according to our data, laboratory tests of similar PowerTune and ZeroCore solutions were carried out several generations of video cards ago. Some of the engineering samples of video cards from AMD series that have long left the market worked exactly like this, completely turning off the cooler when idle.

But it's not just single-GPU users who will benefit from reduced noise and power consumption with AMD's new ZeroCore-enabled graphics cards. Similar improvements await the happy owners of CrossFire systems based on two, three and even four GPUs. It is logical that in the mode of rendering a two-dimensional interface of the operating system, all video cards except the main one should not work at all? But that’s exactly how they work now!

In the case of CrossFire systems on video cards with ZeroCore support in 2D mode, all secondary video cards are put into deep sleep with minimal power consumption and the cooler disabled. This mode works for several single-chip video cards and for dual-chip solutions. In addition, the primary CrossFire graphics card will also enter this mode in case of prolonged inactivity, configured in Windows. The difference in operation looks like this:

By the way, the technology is not as simple as it might seem. AMD engineers had to solve a lot of issues related to the operation of the operating system in idle mode. For example, they found that Windows tries to update information on the screen even when the monitor is turned off. Which, naturally, does not allow you to disable the GPU at all. Therefore, the company's programmers had to take a workaround, ignoring all screen drawing commands when the monitor is turned off in sleep mode.

AMD Eyefinity 2.0 Technology

Naturally, in the new architecture there was also room for improvements to the proven technology for displaying images on multiple monitors - AMD Eyefinity, now in version 2.0. It received new features, higher resolutions, support for more displays and increased flexibility.

This technology is quite interesting, although very few users will find room in the room and muster up the courage to install more than two monitors in their family. But it is better to have the opportunity to always be able to use it than not to have it at all. Moreover, prices for large diagonal monitors are almost not decreasing, but mid-level solutions are constantly becoming cheaper.

Indeed, it is now more profitable to buy three monitors with a screen diagonal of 24″ than one 30-inch. AMD gives just such an example, when a 30″ monitor with a resolution of 2560×1600 costs more than $1000, and three 24″ FullHD ones can be bought for half that price:

But how to spend your money and space in the room is a personal matter for each user. The main thing is that such an opportunity exists. Plus, Eyefinity 2.0 now supports image output in HD3D stereo mode - something that was missing in previous solutions, which were inferior to competing ones in this parameter. Combining AMD Eyefinity and HD3D technologies, the Radeon HD 7970 graphics card is the first single-chip solution to support three monitors operating in stereo mode.

High resolution stereo rendering requires a very fast data interface. And with previous versions of HDMI outputs, the capabilities were limited to 24 Hz per eye, which is quite enough for watching Blu-ray 3D movies, but clearly too low for gamers.

For such tasks, they began to use the frame packing format, when the frames for the left and right eyes are combined into one, and the AMD Radeon HD 7970 supports the HDMI 1.4a frame packing format for stereo image output. This is the first video card to support 3 GHz HDMI with frame packing, when each eye receives a FullHD picture with a frequency of 60 Hz (120 Hz in total):

Another interesting new product seems to us to be the Discrete Digital Multi-Point Audio (DDMA) multi-channel audio output technology, working together with Eyefinity. All previous GPUs are capable of outputting only one audio stream via HDMI and DisplayPort. That is, even if three monitors located in different rooms are connected to the PC via HDMI, only one audio channel is transmitted. But the AMD Radeon HD 7900 received support for simultaneous output of several independent audio channels, which may well be useful in some multi-monitor configurations.

The same feature will be very useful for use in video conferencing with the display of several interlocutors on separate screens, as well as multitasking applications such as playing on three monitors with game audio and viewing news on a separate screen with an independent audio stream. Previously, for all this it was necessary to use several separate audio systems, but now everything works as conveniently as possible.

Eyefinity software support has not been forgotten either; the technology is updated almost every month - new opportunities appear. So, back in October, support for resolutions up to 16384x16384 and new multi-monitor configurations appeared: horizontal and vertical 5x1, as well as based on six monitors in 3x2 mode.

The December update to the AMD Catalyst video driver made it possible to work together between Eyefinity and HD3D, and in February they promise support for custom resolutions, taskbar placement settings and improved preset management.

Output to six monitors can be achieved using two DisplayPort 1.2 ports and two MST hubs (which we wrote about earlier), while three or even four monitors will require only one port and the corresponding hub. Such hubs allow flexible configuration of the image output system; they support up to four FullHD devices per DisplayPort 1.2 connector and should be available for sale by the summer of 2012.

Speaking of resolution. High resolution or even ultra-high - Ultra High Resolution. Current devices with a resolution of 4000 pixels on the larger side require connection using several cables at once: two DP 1.1 or four DVI. Monitors of this resolution of the next generation will be connected via only one cable: DP 1.2 HBR2 or HDMI 1.4a 3 GHz. And the new video card from AMD is already ready for such monitors, again it became the first in the world.

Video encoding and decoding

It’s quite natural that the AMD Radeon HD 7970 includes the same UVD unit for decoding video data, which appeared in the previous generation of the company’s video chips. It simply does not need any modifications, supporting the multi-stream MVC codec, decoding MPEG-2/MPEG-4 (DivX), VC-1 and H.264 formats, as well as decoding two FullHD streams in all supported formats.

AMD solutions provide maximum quality of video stream decoding, use several dozen special quality improvement algorithms and provide maximum results in quality tests like HQV. Among the supported features, we note: color and tone adjustment, noise reduction, sharpening, high-quality scaling, dynamic contrast, advanced deinterlacing, and inverse telecine. Here's an example of improving contrast on the fly:

But decoding on all video chips has been more or less in order for a long time. All new GPUs provide decent quality and performance when viewing video data. But video encoding on the GPU is still in its infancy and the main complaints from users are aimed at the low quality of the resulting compressed image.

Perhaps the new Radeon HD 7000 series can help with this, because all GPUs in the series include a Video Codec Engine (VCE) video encoding unit. The Radeon HD 7970 model became the first video card to support hardware-accelerated video encoding and compression using a specialized unit (previously, stream processors took part in encoding).

Quality and performance should be clearly better than before, with support for 1080p encoding at 60 frames per second, and even faster than real time. It’s difficult to say anything about quality without tests, but we are promised different levels of encoder optimization for video data and games, as well as variable compression quality (the ability to choose between increasing quality or performance).

There is currently no place to try VCE - there are simply no applications that support it, but AMD is working with partners such as ArcSoft to provide support for VCE in their respective software products. In the future, we plan to release a software library to accelerate video encoding, which will make it easier for developers to support the next generation of AMD products.

Encoding can be performed in two modes: full and hybrid (using the capabilities of GPU stream processors). Full mode is designed for tasks that require maximum energy efficiency and consistent performance levels. Full mode encoding on VCE is faster than real time and provides low latency. But there is also a hybrid mode:

In this mode, GPU mathematical blocks also work together with VCE. All of the highly parallelizable stages, which are outlined in yellow in the diagram, can harness the power of GCN compute units, while a dedicated VCE unit handles efficient hardware entropy encoding. This mode is well suited for video cards with great mathematical power, such as the Radeon HD 7970. Questions remain about the quality of these two modes, but this requires careful analysis in a separate article.

AMD Steady Video

In addition to encoding and decoding video data, there is another area where the power of new graphics from AMD can be used - improving poor quality videos taken handheld, without the use of a tripod or other similar image stabilization means. The video stabilization technology is called AMD Steady Video, and its second version has already been released.

The operating algorithm of the software stabilizer is quite simple: based on the video stream, statistics about camera movement (shift, rotation, zoom) are collected and this movement is compensated in the current frame, relative to the previous ones - the image is shifted, rotated and scaled so that the picture does not jump much and remains stable.

As simple as it sounds, it is just as difficult to implement. Simply because there are two million pixels on the screen, and up to 30 or even 60 frames per second. Imagine how many calculations need to be done to track all possible frame displacements. We have already written above about the QSAD function used in video processing; it is also used in Steady Video 2.0 to speed up the motion detection algorithm. So the GPU must process random shifts with an amplitude of up to 32 pixels in any direction and this requires performance corresponding to more than 500 billion SAD operations per second (for 1920x1080 at 60 FPS).

Due to the support of new QSAD instructions in the Radeon HD 7970, its advantage over powerful CPUs in the motion detection algorithm exceeds 10x! That is, we will now be provided with high-quality video, and not only when processing home videos in video editors, but also when watching other people’s online videos, shot with who knows what and who knows how.

Details: Radeon HD 7800 series

Chip codename: "Pitcairn"
Manufacturing technology: 28 nm
2.8 billion transistors (slightly more than the Cayman, which is the basis of the Radeon HD 6900 series)
A unified architecture with an array of common processors for stream processing of numerous types of data: vertices, pixels, etc.
Hardware support for DirectX 11.1, including Shader Model 5.0
256-bit memory bus: four 64-bit wide controllers supporting GDDR5 memory
Core frequency: up to 1000 MHz (for Radeon HD 7870)
20 GCN compute units, including 80 SIMD cores, consisting of a total of 1280 ALUs for floating point calculations (integer and floating formats, support for FP32 and FP64 precision within the IEEE 754 standard)
80 texture units, with support for trilinear and anisotropic filtering for all texture formats
32 ROP units with support for anti-aliasing modes with programmable sampling of more than 16 samples per pixel, including with FP16 or FP32 frame buffer format. Peak performance up to 32 samples per clock, and in Z only mode - 128 samples per clock

Radeon HD 7870 graphics card specifications

Core frequency: 1000 MHz
Number of universal processors: 1280
Number of texture blocks: 80, blending blocks: 32
Memory type: GDDR5
Memory capacity: 2 gigabytes
Theoretical maximum fill rate: 32.0 gigapixels per second.
Theoretical texture sampling rate: 80.0 gigatexels per second.
One CrossFire connector
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, two Mini-DisplayPort 1.2
Power consumption: from 3 to 175 W
Two 6-pin power connectors
Dual slot design
Recommended price for the US market: $349

Radeon HD 7850 graphics card specifications

Core clock: 860 MHz
Number of universal processors: 1024
Number of texture blocks: 64, blending blocks: 32
Effective memory frequency: 4800 MHz (4x1200 MHz)
Memory type: GDDR5
Memory capacity: 2 gigabytes
Memory bandwidth: 153.6 gigabytes per second.
Theoretical maximum fill rate: 27.5 gigapixels per second.
Theoretical texture sampling rate: 55.0 gigatexels per second.
One CrossFire connector
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, two Mini-DisplayPort 1.2
Power consumption: from 3 to 130 W
Dual slot design
US MSRP: $249

And this time the principle of naming the company’s products was not changed and the trends of the previous series were continued. The mid-budget series of video cards based on the GCN architecture differs from the top and budget lines by the second digit in the index: instead of 7 and 9, the number 8 is placed, which is quite logical. Since AMD took the psychological threshold of 1000 MHz for the GPU frequency, the Radeon HD 7870 received the addition of “GHz Edition” to the name, indicating that this frequency was taken.

From the name it is clear that the Radeon HD 7800 is more productive than the HD 7700, but has a lower speed compared to the older models - the HD 7900. As for comparison with NVIDIA solutions, the older released model HD 7870 at the time of release competes with the GeForce GTX video card 570, and the younger one is aimed at fighting the GTX 560 Ti, and NVIDIA has not yet released new 28 nm mid-range chips.

Both models of AMD video cards have GDDR5 memory of the same capacity of 2 gigabytes. They both use a 256-bit memory bus, and so could be configured with 1, 2, or 4 GB. 1 GB is too little, and 4 GB is too expensive for this price segment. Therefore, we can say that the ideal volume of 2 GB of video memory was chosen, quite sufficient for the vast majority of games, even in high resolutions, and not too expensive in terms of cost.

Otherwise, from a consumer point of view, the HD 7850 and HD 7870 models are still different. The older Radeon HD 7870 has higher power consumption, so it needs two additional 6-pin power connectors, while the HD 7850 is content with only one of them. Both boards have a dual-slot cooling system design, but most manufacturers produce boards with their own design for at least a cooler, or even a printed circuit board.

Architectural features of the Radeon HD 7800 family

Above in the text, we carefully described all the features of the new Graphics Core Next (GCN) architecture, so we will repeat only the most important ones. All of the company's new GPUs offer excellent capabilities and performance not only in graphics processing, but also in non-graphics computing, including a mixture of different types of calculations. Also, the new GCN architecture offers a significant simplification of code optimization tasks, simplified development and support, as well as stable and predictable performance and, in general, quite high efficiency.

The base block of the new architecture is the GCN block, and all Southern Islands series GPUs are assembled from them. Consider the block diagram of the Pitcairn chip:

The diagram shows the Radeon HD 7870 graphics processor (the “simplified” HD 7850 differs from it in several disabled units); we see 20 computing units of the GCN architecture. In the case of the junior solution of the Radeon HD 7800 series, four of them were disabled, and the number of active blocks in it is 16. This corresponds to 1280 and 1024 stream processors, respectively (exactly as in the case of the HD 7700 family, only there are exactly twice as many blocks) . Since each GCN block contains four texture units, the final number of TMUs for the older model is 80 TMUs, and for the younger one - 64 TMUs.

But the number of ROP units and memory controllers in the HD 7870 and HD 7850 is also the same, as in the solutions of the youngest line. The number of ROP blocks was left quite high - 32 pieces for both models. The memory bus of Pitcairn-based boards is reduced to 256-bit; it is assembled from four 64-bit channels. This is not bad for a solution of this level, although it is one and a half times less than in the top line, because the memory bus is traditionally the first thing to be cut. It’s good that the use of fast GDDR5 memory gave a relatively high bandwidth of 153 GB/s.

Like other GCN architecture chips, Pitcairn incorporates a 9th generation tessellator unit, which features numerous buffering and caching optimizations to significantly improve geometry processing performance. Here is a comparison of the new AMD board with the previous generation solution in a synthetic problem, which suggests an increase in tessellation speed of up to four times:

Likewise, many AMD technologies are supported, which were introduced and improved in the new video chips of the Radeon HD 7000 line. Here is an incomplete list: PowerTune, ZeroCore, Eyefinity 2.0, HD3D, Steady Video, improvements in the quality of texture filtering, etc. All this is written in more detail above. Let's add to the list that the Radeon HD 7800 fully supports both the improved anti-aliasing algorithm MLAA 2.0 and super-sampling anti-aliasing (SSAA).

When it comes to comparing gaming performance, the Radeon HD 7870 is significantly faster than its direct competitor, the GeForce GTX 570, especially given the lack of 1.25 GB of video memory in the latter (compared to 2 GB in the solutions under review), observed in modern games at high rendering resolutions . The younger Radeon HD 7850 can be compared with the GeForce GTX 560 Ti, and here it can no longer boast of memory capacity. However, according to AMD's measurements, their new solution is still faster than its competitor in most games.

Details: Radeon HD 7700 series

Chip codename: "Cape Verde"
Manufacturing technology: 28 nm
1.5 billion transistors (less than Barts, which is the basis of the Radeon HD 6800 series)
A unified architecture with an array of common processors for stream processing of numerous types of data: vertices, pixels, etc.
Hardware support for DirectX 11.1, including Shader Model 5.0
Core frequency: up to 1000 MHz (for Radeon HD 7770)
10 GCN compute units, including 40 SIMD cores, consisting of a total of 640 floating point ALUs (integer and floating point formats, support for FP32 and FP64 precision within the IEEE 754 standard)
40 texture units, with support for trilinear and anisotropic filtering for all texture formats
Integrated support for up to six monitors, including HDMI 1.4a and DisplayPort 1.2

Radeon HD 7770 graphics card specifications

Core frequency: 1000 MHz
Number of universal processors: 640
Number of texture blocks: 40, blending blocks: 16
Memory type: GDDR5
Memory capacity: 1 gigabyte
Theoretical texture sampling rate: 40.0 gigatexels per second.
One CrossFire connector
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, two Mini-DisplayPort 1.2
Power consumption: from 3 to 80 W
One 6-pin power connector
Dual slot design
Recommended price for the US market: $159

Radeon HD 7750 graphics card specifications

Core frequency: 800 MHz
Number of universal processors: 512
Number of texture blocks: 32, blending blocks: 16
Effective memory frequency: 4500 MHz (4x1125 MHz)
Memory type: GDDR5
Memory capacity: 1 gigabyte
Memory bandwidth: 72 gigabytes per second.
Theoretical maximum fill rate: 12.8 gigapixels per second.
Theoretical texture sampling rate: 25.6 gigatexels per second.
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, one DisplayPort 1.2
Power consumption: from 3 to 55 W
Does not require additional power
Single slot design
Recommended price for the US market: $109

An inexpensive series of video cards based on the GCN architecture differs from the top and mid-range lines by the second digit in the index: the number 9 is occupied by the number 7, as was the case before. The Radeon HD 7770 is a more productive solution, but there is also a younger model - the HD 7750. The older board at the time of its release had no direct competitors on the market, located somewhere between the GeForce GTX 560 and GTX 550 Ti, and the younger one is aimed at fighting the GTX 550 Ti. For the HD 7770, a competitor was later announced in the form of the GeForce GTX 560 SE (all NVIDIA solutions are based on older GPUs).

Both models of AMD video cards under consideration have GDDR5 memory of the same capacity of 1 gigabyte. Due to the use of a 128-bit memory bus, they could be equipped with 2 GB, but such an amount of GDDR5 memory would be too expensive for their price segment. Therefore, so far models with this volume have been released, although in the future it is possible that variants with 2 GB of video memory will be released. For now, they decided to leave this capacity for the HD 7800.

In terms of other consumer characteristics, the HD 7750 and HD 7770 models are quite different. If the older Radeon HD 7770 has a two-slot cooling system design and its cooler is covered with a plastic casing like older solutions, then the younger HD 7750 looks noticeably simpler, occupying one slot and having a simple cooler. However, most manufacturers still produce boards with their own designs. The power consumption of new models in this price range is also different, the older one has one 6-pin additional power connector, and the younger one uses power received via PCI Express.

Architectural features of the Radeon HD 7700

The basic block of the new architecture is the GCN block, and all GPUs in the series are assembled from them. Each of the available GCN blocks is capable of scheduling and distributing commands itself, and one computing block can execute up to 32 independent command threads. Let's look at the block diagram of the Cape Verde chip:

The diagram shows the Radeon HD 7770 graphics processor (the “stripped-down” HD 7750 has several disabled units); we see 10 computing units of the GCN architecture. In the case of the junior solution of the Radeon HD 7700 series, it was decided to disable two of them, and the number of blocks became 8. This corresponds to 640 and 512 stream processors. And since each GCN block contains 4 texture units, the final number of TMUs for the older model is 40 TMUs, and for the younger one - 32 TMUs.

The number of ROP units and memory controllers in the HD 7770 and HD 7750 is no different, and they decided not to cut the ROP too much, leaving 16 of them each. But the Cape Verde's memory bus is cut to 128-bit, which is assembled from two 64-bit channels. In general, this is three times less than in the top series, and we saw another confirmation that the memory bus is traditionally the first thing to be cut in inexpensive chips. Although the use of fast GDDR5 memory made it possible to maintain a relatively high (for such inexpensive solutions) bandwidth of 72 GB/s.

All that remains for us to note is the rather large amount of second-level cache - as much as 512 kilobytes (compare with 768 KB for the top chip - apparently the L2 cache does not take up too much space on the chip), as well as improvements in geometric performance. Like the top-end chip, the Cape Verde features a 9th-generation tessellator featuring numerous buffering and caching optimizations to significantly improve geometry processing performance compared to the Radeon HD 6000 series.

In general, we will not repeat all the information about AMD technologies that have been introduced and improved in the new video chips of the Radeon HD 7000 line (here is an incomplete list: PowerTune, ZeroCore, Eyefinity 2.0, HD3D, Steady Video, improvements in the quality of texture filtering, etc. .p.), all this is written in detail above. The HD 7700 series supports all the features listed there, including AMD Eyefinity 2.0 with six monitors and stereo rendering, and also has an improved video decoding and encoding unit.

But what about the most important thing - gaming performance? The first estimates of rendering speed can always be made from the manufacturer’s presentations. AMD believes that the Radeon HD 7770 is located somewhere in the middle between the GeForce GTX 560 and GeForce GTX 550 Ti, respectively, and compares it in its materials with the second model of the competitor.

But they do not compare the Radeon HD 7750 with anything, simply noting that most modern games are playable on this model at maximum settings in FullHD resolution. However, this is not surprising, since in recent years there have been practically no PC exclusives, and multi-platform games are much less demanding. So the Radeon HD 7700 series boards are perfect for undemanding users.

Details: Radeon HD 7790 model

Chip codename: "Bonaire"
Manufacturing technology: 28 nm
2.08 billion transistors (more than the Cape Verde in the Radeon HD 7700, but less than the Pitcairn in the Radeon HD 7800)
A unified architecture with an array of common processors for stream processing of numerous types of data: vertices, pixels, etc.
Hardware support for DirectX 11.1, including Shader Model 5.0
128-bit memory bus: two 64-bit wide controllers supporting GDDR5 memory
Core frequency: 1000 MHz
14 GCN compute units, including 56 SIMD cores, consisting of a total of 896 floating point ALUs (integer and floating point formats, support for FP32 and FP64 precision within the IEEE 754 standard)
56 texture units, with support for trilinear and anisotropic filtering for all texture formats
16 ROP blocks with support for antialiasing modes with the ability to programmably sample more than 16 samples per pixel, including with FP16 or FP32 frame buffer format. Peak performance up to 16 samples per clock, and in Z only mode - 64 samples per clock

Radeon HD 7790 graphics card specifications

Core frequency: 1000 MHz
Number of universal processors: 896
Number of texture blocks: 56, blending blocks: 16
Memory type: GDDR5
Memory capacity: 1 gigabyte
Memory bandwidth: 96 gigabytes per second.
Theoretical maximum fill rate: 16.0 gigapixels per second.
Theoretical texture sampling rate: 56.0 gigatexels per second.
One CrossFire connector
PCI Express 3.0 bus
Connectors: DVI Dual Link, HDMI 1.4, two Mini-DisplayPort 1.2
Power consumption: from 3 to 85 W
One 6-pin power connector
Dual slot design
US MSRP: $149

The inexpensive video card model, based on a new mid-budget chip, differs from the previous top model of the HD 7700 subfamily by the third digit in the index: instead of 7, they put the number 9, which indicates an increase in performance. At the same time, the Radeon HD 7790 index clearly indicates that this is a less powerful video card compared to the line one step higher - HD 7800.

However, everything is not so simple here either - it can certainly compete with the younger HD 7850. But the recommended price of the Radeon HD 7790 is set at $149, that is, approximately in the middle between the prices of the HD 7770 and HD 7850. As for the solutions of a competitor from the same price segment, the release of the HD 7790 was clearly intended to have something to fight with NVIDIA GeForce GTX 650 Ti, based on the GK106 chip, which is located exactly between the HD 7770 and HD 7850 in price and speed. But NVIDIA immediately responded to the release of the new board by AMD, releasing to the market an overclocked version of the GeForce GTX 650 Ti Boost, which has greater performance.

This AMD video card model has GDDR5 memory with a capacity of only 1 gigabyte. The GPU has a 128-bit memory bus, and theoretically it would be possible to supply 2 GB, but this amount of fast GDDR5 memory is still too expensive for this price segment, and AMD released a model with a smaller capacity, although it may not be enough for some modern games even at not the highest settings and resolution. However, it is possible to release video cards from partners with 2 GB of video memory.

Like the models next to it in the line, the Radeon HD 7790 has a dual-slot cooling system design, which is covered with a plastic casing. Although most manufacturers still release motherboards with their own cooler design, so the reference one is not so important. Interestingly, the power consumption of the new model has not increased too much compared to the HD 7770, but the improvement in energy efficiency was expected. By the way, this is why the new product also has only one 6-pin additional power connector.

Architectural features

The new Bonaire graphics processor, on which the released Radeon HD 7790 model is based, belongs to the same Graphics Core Next (GCN) architecture that we have been familiar with for a year and a half, but AMD calls it GCN 1.1, hinting at small changes. In fact, the chip is practically no different architecturally from the previous ones, although there are indeed some minor changes. For example, the new architecture introduced instructions useful for heterogeneous system architecture (HSA), support for a larger number of simultaneously executed threads, as well as a new version of AMD PowerTune technology, which we will talk about later. But all these changes cannot be called significant, because there is nothing new in the basic blocks and improving their efficiency.

Therefore, you can safely refer to, which carefully describes all the features of the new Graphics Core Next (GCN) architecture, and here we will only repeat the most important characteristics and features of a particular product. All of AMD's latest GPUs offer excellent capabilities and performance in both graphics and non-graphics computing, and a mixture of the two. The new GCN architecture has also greatly simplified optimization and software development tasks while maintaining high efficiency.

As you know, the basic block of the architecture is the GCN block, from which all Southern Islands series graphics processors are assembled. The GCN computational block is divided into subsections, each of which works on its own command flow. GCN blocks have 64 KB of dedicated local data storage for exchanging data or expanding the local register stack. The block also has a first-level cache with read-write capabilities and a full-fledged texture pipeline with sampling and filtering units. Each of the existing GCN blocks is capable of scheduling and distributing commands itself, and one computing block can execute several independent command streams. Let's look at the block diagram of the new chip:

Bonaire's design confirms the new solution's goal of offering performance between Cape Verde, which has 10 GCN compute units, and Pitcairn, with its 20 GCN units. These two GPUs, released in 2012, are almost exactly half the size of each other, leaving a fairly large performance gap in the middle that Bonaire has now filled.

The diagram shows a graphics processor in the form of a Radeon HD 7790, which is a complete solution without cutting down any blocks. The chip includes 14 computing units of the GCN architecture, which corresponds to 896 stream processors. Since each GCN block contains 4 texture units, the final number of TMUs for the new model is 56 TMUs. That is, Bonaire is exactly 1.4 times faster than the Cape Verde chip in terms of the speed of mathematical calculations and texture fetches, assuming equal frequencies.

But the number of ROP units and memory controllers in Bonaire and Radeon HD 7790 is similar to what we saw in Cape Verde and Radeon HD 7770 - they decided to leave 16 ROP units, and the memory bus of the new chip is 128-bit, assembled from two 64-bit bit channels. The small number of ROP blocks may be the “Achilles heel” of the solution, since the use of fast GDDR5 memory made it possible to provide a relatively high bandwidth of 96 GB/s, but nothing can be done about ROP performance.

But the new GPU has improvements in geometric performance and tessellation speed. Yes, Cape Verde also has a 9th generation tessellator, but Bonaire also doubled the number of geometric blocks, rasterizers and command processors (indicated as ACEs in the diagram) - now there are two of them all. This improvement gives Bonaire the ability to process up to two geometric primitives per clock cycle - just like the more powerful Pitcairn and Tahiti.

As you remember, it was in the Radeon HD 7770 that AMD first reached the important psychological milestone of the GPU clock frequency of 1 GHz. So, the HD 7790 also has exactly the same reference frequency of 1 GHz, so the increase in performance compared to the HD 7770 will be justified solely by architectural changes and an increase in the number of execution units.

But the video memory operating frequency of the new product is much higher. If the HD 7770 had a relatively low memory frequency of 4.5 GHz, then the HD 7790 is equipped with fast GDDR5 memory operating at 6 GHz, which provides a third more bandwidth. The video memory bandwidth increased by 33% compared to models of the Radeon HD 7700 subfamily led to a clear increase in gaming performance. AMD provides this chart comparing frame rates on the HD 7790 with memory running at 4.5 and 6.0 GHz:

The maximum acceleration from an increase in memory bandwidth was achieved in games such as StarCraft II and Crysis 2. And on average, a 33% increase in memory bandwidth gives somewhere around a 10% increase in the average frame rate in a set of modern games. This is a good indicator, showing that memory bandwidth is quite important nowadays, although it is not the only focus on productivity. Although it is quite possible that with more ROP, Bonaire's speed would be even higher...

It is clear that the average power consumption has increased slightly compared to the HD 7770. If for the old model this value is 80 W, then for the HD 7790 it is 85 W - this is a very small price to pay for a theoretical performance increase of 33-40%! Architectural improvements (PowerTune), designing a new GPU using experience from previous ones, as well as continuous improvement of the technical process at TSMC - all this led to a slight increase in consumption with a significant improvement in speed characteristics.

As for the chip area and the number of transistors in Bonaire, the new chip is clearly larger than Cape Verde, but the addition of computational, texture and geometry units could not go unnoticed. According to these parameters, Bonaire is also located approximately in the middle between Cape Verde and Pitcairn. Bonaire contains 2.08 billion transistors in a 160 mm 2 chip, for Cape Verde these figures are 1.5 billion and 123 mm 2, respectively, and for Pitcairn - 2.8 billion transistors and 212 mm 2 chip area.

Naturally, the new chip supports all AMD technologies that were introduced and improved in the new Radeon HD 7000 family (an incomplete list: PowerTune, ZeroCore, Eyefinity, HD3D, Steady Video, improved texture filtering quality, etc.), both all this is written in detail in the article AMD Radeon HD 7970: The New Single-Processor Leader. The HD 7790 supports all the features listed there, including AMD Eyefinity 2.0 with six monitors and stereo rendering, and also has an improved video decoding and encoding unit.

Improved PowerTune technology

Back in 2010, AMD introduced PowerTune technology in its Cayman chip (AMD Radeon HD 6900 series). This GPU was the first to feature dynamic power management, called PowerTune. It allowed higher maximum clock speeds for typical applications, while avoiding too much power consumption in specialized stability tests like FurMark. Then the technology was applied to the dual-chip AMD Radeon HD 6990 model, which needed it even more for obvious reasons.

The technology received a serious update in mid-2012, when automatic frequency increase - Boost - was added to AMD PowerTune. In the AMD Radeon HD 7970 GHz Edition, this algorithm made it possible to further improve performance compared to the regular version of the video card. The PowerTune operating algorithm in video cards without automatic overclocking uses three states: idle, low-3D and full-speed. The HD 7970 GHz also added a Boost overclocking mode. PowerTune serves to stay within the required consumption, switching to a lower load mode when necessary. In this case, the technology sharply reduces the clock frequency. In practice, such jumps are rare - due to the large gap between the two active modes.

Reducing the GPU clock speed reduces power consumption, but for more efficient control, you also need to reduce the voltage. That's exactly what the Radeon HD 7790 does. The new Bonaire graphics chip has eight states with different frequencies and voltages, allowing you to achieve higher clock speeds than before, while the GPU is always running at the optimal voltage and frequency. Switching between states is based on GPU load, as well as the current energy consumption of the video chip.

In the new algorithm, PowerTune does not have to sharply reset the frequency when the consumption level is exceeded, and along with the frequency, the voltage also decreases. Transitions between states must be as fast as possible so as not to exceed the consumption limit even for a short time, so Bonaire switches PowerTune states every 10 ms, that is, the state of the chip changes 100 times every second.

With such a constant change in frequencies, third-party applications like MSI Afterburner and GPU-Z will not show instantaneous clock speed values, but average ones over a certain period of time - the so-called “effective” frequency. Another interesting development is that AMD is opening up new PowerTune settings to third-party applications. Partners can also set their own PowerTune settings, which will help when creating factory overclocked video card models and provide more options that are not limited by AMD reference values. True, different PowerTune settings can lead to the fact that video cards of the same model from different manufacturers will not only have different clock frequencies, but also the algorithm for changing them over time, which will make it difficult to compare under the same conditions.

Sales of Radeon HD 7790 video cards on the market began at the very beginning of April 2013. AMD, together with its partners, organized the release of both motherboards with reference frequencies and factory overclocked solutions. And now both manufacturers are launching new video cards to the market in approximately the same way, with various options quickly available from their partners. In fact, the partners released almost more overclocked versions of the HD 7790 than regular ones, and the graphics chips in them operate at frequencies of about 1075 MHz.

Details: Radeon HD 7990 model

Codename "Malta"
Manufacturing technology: 28 nm
2 chips with 4.3 billion transistors each
A unified architecture with an array of common processors for stream processing of numerous types of data: vertices, pixels, etc.
Hardware support for DirectX 11.1, including Shader Model 5.0
Dual 384-bit memory bus: twice six 64-bit wide controllers with support for GDDR5 memory
GPU frequency: 1000 MHz
Twice 32 GCN compute units, each with 128 SIMD cores, consisting of a total of 4096 floating point ALUs (integer and floating point formats, support for FP32 and FP64 precision within the IEEE 754 standard)
2x128 texture units, with support for trilinear and anisotropic filtering for all texture formats
2x32 ROP units with support for antialiasing modes with the ability to programmably sample more than 16 samples per pixel, including with FP16 or FP32 frame buffer format. Peak performance up to 64 samples per clock, and in Z only mode - 256 samples per clock
Integrated support for up to six monitors via HDMI 1.4a and DisplayPort 1.2 interfaces

Radeon HD 7990 Graphics Card Specifications

Core frequency: 1000 MHz
Number of universal processors: 4096
Number of texture blocks: 2x128, blending blocks: 2x32
Effective memory frequency: 6000 MHz (4x1500 MHz)
Memory type: GDDR5
Memory capacity: 2x3 gigabytes
Memory bandwidth: 2x288 gigabytes per second.
Theoretical maximum fill rate: 64 gigapixels per second.
Theoretical texture sampling rate: 256 gigatexels per second.
One CrossFire connector
PCI Express 3.0 bus
Connectors: DVI Dual Link, four Mini-DisplayPort 1.2
Power consumption up to 375 W
Two 8-pin auxiliary power connectors
Dual slot design
Recommended price for Russia - 32,999 rubles. (for the USA - $999).

Already in the second generation of AMD video cards, the naming principle for dual-chip models remains unchanged. The top solution on two powerful video chips differs from the previous generation model corresponding to its class in the first digit in the index: instead of 6, it received the number 7, indicating a new series. The announced video card differs from the single-chip solution in the third digit, indicating maximum performance within the generation.

As for comparison with competitors, the main rival for the Radeon HD 7990 model announced today is the GeForce GTX 690 video card, released almost a year ago, and it is these dual-chip solutions that will have to fight with each other. True, NVIDIA also has another powerful solution, but based on a single GPU - GeForce GTX Titan, which can also be considered a competitor to the board in question from AMD.

The new dual-chip Radeon video card is equipped with 3 gigabytes of GDDR5 memory for each GPU, which is due to the 384-bit memory bus of Tahiti chips. This volume is quite justified for a product of such a high level, since in some modern gaming applications, with maximum settings, anti-aliasing enabled and high resolutions, a smaller amount of memory (2 gigabytes per chip or less) may no longer be enough. And even more so when rendering in stereo mode or on multiple monitors in Eyefinity mode.

It is clear that such a powerful dual-chip video card has a massive dual-slot cooling system, which differs from traditional coolers for AMD cards. It features a massive radiator hidden under a casing with three large fans operating at relatively low speeds. The power consumption of a card with two GPUs on board is quite high for obvious reasons, and it has two 8-pin power connectors, but at least this is not three, as was the case in non-reference samples based on two Tahiti chips.

Architecture

Since the video card, codenamed "Malta", is based on two "Tahiti" GPUs from the Southern Islands family, you can simply refer to, which carefully describes all the features of the current Graphics Core Next (GCN) architecture. In basic materials we repeat only the most important characteristics and features of specific products.

The basic block of the architecture is the GCN block, from which all GPUs in the series are assembled. The compute unit is divided into subsections, each of which works on its own instruction flow, it has dedicated local data storage, a read-write L1 cache, and a full texture pipeline with fetch and filter units. Each of the GCN blocks is capable of scheduling and distributing commands itself, and one computing block can execute several independent command streams. The Radeon HD 7990 uses two Tahiti chips already known to us:

The diagram of the graphics processor (there are two of these in the Radeon HD 7990) shows 32 computing units of the GCN architecture and all of them are active. Previously, it was assumed that for a two-chip solution it would be necessary to disable some of them, and even lower the frequency in order to enter the power consumption limits of 375 W, but AMD engineers managed to successfully solve this difficult problem. Perhaps a special new revision of Tahiti with reduced power consumption was released, or the chips simply undergo a very strict selection process.

Since each GCN unit contains 16 texture units, the number of TMUs is 128 units per chip, which gives a final performance of 256 gigatexels per second, which is very good for a competitor to the GeForce GTX 690. Number of ROP units and memory controllers in HD 7990 also did not change compared to its single-chip counterpart; they were left at 32 and 6 pieces per GPU, respectively. The Radeon HD 7990 has dual 384-bit memory buses made up of twelve 64-bit channels, which provides a total memory bandwidth of 576 GB/s - another record-breaking figure.

Otherwise, the new board supports all modern AMD technologies that were introduced and improved in the new video chips of the Radeon HD 7000 line: PowerTune, ZeroCore, Eyefinity 2.0, HD3D, Steady Video, improved texture filtering quality, etc. All this is written in detail above in the description of the Radeon HD 7970, and there is simply no point in repeating it.

Cooling system and power consumption

In the case of such serious dual-chip boards, a highly efficient cooling system becomes especially important. If in the case of solutions from partners based on two Tahiti, three-slot solutions were used, and in the case of ASUS ARES II, water cooling was used, in this case it was necessary to make do with less effort, so a cooler was designed with a very massive radiator and three fans with improved acoustic characteristics.

The noise level of the cooling system and the temperature provided for GPUs are among the most important consumer characteristics for any video card, including a top-end solution intended for enthusiasts. A cooling system that is too loud or ineffective will be considered by buyers as a less profitable purchase, all other factors being (roughly) equal. So AMD took this issue very seriously in the case of the Radeon HD 7990 model, when compared with other top solutions on the market. Let's look at the acoustic characteristics of the new system:

The diagram shows the noise level from three different video cards: the Radeon HD 7990 and two competitors: the dual-chip GeForce GTX 690 and the single-chip GTX Titan from NVIDIA. Moreover, the noise was measured under different conditions - in idle mode (System Idle) and at maximum load using Furmark. If you believe the figures provided by AMD, then even the single-chip Titan does not reach their new product in terms of noise level from the cooler, not to mention the dual-chip GTX 690, which is the loudest in this comparison.

But was such impressive acoustic performance achieved at the expense of GPU temperature? The following chart shows GPU temperatures measured on AMD's Radeon HD 7990 and the same two competitors. This time, AMD specialists used only the high load mode when testing at Furmark.

And again, a “tricky” coordinate axis is used with the origin not at zero. The real difference between 80 and 82 degrees for the Radeon HD 7990 and GTX Titan will be virtually imperceptible, although 87 degrees for the GTX 690 clearly stands out for the worse. Again, we note that all of these tests were conducted by a stakeholder and are subject to independent verification.

In terms of power consumption, there is nothing new in the dual-chip solution, but there is also support for the previously announced ZeroCore Power technology. This technology helps achieve significantly lower power consumption in deep idle (or sleep) mode with the display device turned off. In this mode, the idle GPU is almost completely turned off and consumes less than 5% of the energy of the full mode, disabling most functional units. And in the case of a two-chip board, it is even more important that in the CrossFire system, when rendering a two-dimensional interface of the operating system, all GPUs except the main one do not work at all. That is, in the case of the Radeon HD 7990, one of the chips in 2D mode will be immersed in deep sleep with minimal energy consumption, and the second one can “fall asleep” in deep PC idle mode.

Modern games every year require more and more powerful video cards for graphics processing. One of the budget solutions for gamers will be the AMD Radeon HD 7800 Series. Let's look at the technical characteristics of this series, as well as its features and performance in games.

Let's look at the characteristics of the AMD Radeon HD 7800 Series in the form of a table:

Technical process	28 nm
GPU	Pitcairn
GPU	Frequency (min. max. on models)	800-1000 MHz
RAM	Type	GDDR5
	Volume	2 GB
	Frequency	800-1200 MHz
	Bandwidth	153.6 GB/s
Interfaces	Memory size	256-bit
Interfaces	Tire type	PCI Express 3.0
Architecture	GCN
	Stream processors per block	from 64 to 80
	Number of blocks	from 16 to 20
	Total stream processors	1024-1280
	Cores for geometry processing	2 pcs.
	Cores for asynchronous computing	2 pcs.
Connectors	HDMI DisplayPort 1.2
Supported technologies and software	DirectX 11	Yes
	OpenGL 4.2	Yes
	Eyefinity (monitor merging)	up to 6 pcs.
	ZeroCore Power	Sleeping mode
	Catalyst	Branded Drivers and setup
	App Acceleration	Improving video playback quality
	AMDHD3D	3D graphics processing
	Power Tune	Dynamically adjusting power consumption

The line was produced in March 2012. Based on it, the following models were released:

HD7850;
HD7870;
HD7890.

At the moment, the model range is no longer produced. At the start of sales, the average price in stores was $249 and $349.

Review

At AMD, after entering a new technological process, it was decided to divide the overall production series into subgroups. Therefore, a total of 4 lines were formed based on the 28nm process technology, which are presented in the table:

The HD 7800 with the "Pitcairn" GPU uses the Graphic Core Next microarchitecture. The series was launched in March 2012 and is no longer in production today.

At one time, video cards from Pitcairn were quite popular and showed an excellent price/quality ratio. As of 2018, the current series is not popular and it is extremely difficult to find a device in new condition. Despite the fact that the graphics core is already outdated, when combined with a powerful processor, a PC can run various games at medium and high settings.

What games will run on the AMD Radeon HD 7800 Series

Video cards were released back in 2012, but they can still be used on modern toys. Tests in AMD Radeon HD 7800 Series games were carried out with the following hardware:

Processor: Core I5 6500 3.2 GHz.
RAM: 16 GB DDR4 2133 Dual.
Hard drive: Hitachi 1TB.
Motherboard: Asus H170M-Plus.
Resolution: 1920x1080px.

The results are as follows:

Name of the game	Graphics quality
Assassins Creed Syndicate	Hight	31
WarThunder	Cinema (Ultra)	55-65
Quantum Break	Average	30-42
Assassins Creed Unity	Average	30
Shadow Warrior 2	High	35-45
Dying Light	High	40-50
Fallout 4	Ultra	38-43
GTA 5	Above average	45-50
DOOM	High	40
Rise Of The Tomb Raider	High	30-40
Warface	High	90-100
The Witcher 3: Blood and Wine	High	25-35
World Of Tanks	High	60-80

Overall performance largely depends on the correct combination of processor and video card. If you take a powerful modern generation processor like Ryzen or Core I5, then they will be able to show high FPS in most modern games, even with an old video card.

After analyzing the technical characteristics and tests in games, we come to the following conclusions: it is not recommended to buy for powerful games in 2018; it is better to choose newer models.

The performance will be enough for comfortable home work and for running multiplayer games like CS:GO, World Of Tanks.

How to overclock a video card

To achieve maximum performance, you can overclock the AMD Radeon HD 7800. To do this, you will need to install the driver and configure it.

The main changes need to be made in the “Games” section. If you use a modern driver, profiles can be configured individually for each video game.

Go to the desired profile and go to the “Frame Rate Control” setting. By default, the video card squeezes out maximum FPS and spends all resources on it.

For a comfortable game of shooters, 60 frames per second is enough. For CS:GO, WarFace, WarThunder, it is enough to set the limit to 70 FPS.

Setting OverDrive allows you to adjust operating parameters: GPU and memory frequencies, fan efficiency and power consumption levels. These parameters must be configured individually for each PC build.

Download drivers for AMD Radeon HD 7800 Series

To download drivers for Radeon HD 7800 Series, use the proprietary auto-search program. You can download it on the manufacturer’s official website. There you can also find drivers for each OS version: Windows 7, Windows 10, etc.

Page 2 of 5

"Southern Islands"

First, a little about AMD's labeling of its latest products. The manufacturer has divided them into three levels according to performance. The codename "Cape Verde" refers to the Radeon HD 7700. The name "Pitcairn" hides today's test participants Radeon HD 7870 and HD 7850. High-performance products are called "Tahiti" or Radeon HD 7900. This is shown more clearly below.

Entry level = Cape Verde = Radeon HD 7700 series;
Mainstream = Pitcairn = Radeon HD 7800 series;
High Performance Products = Tahiti = Radeon HD 7900 series.

That is, at the moment AMD has covered all market segments with its 28 nm graphics chips. Only the release of a dual-core video card based on Tahiti chips is expected. Preliminary name Radeon HD 7990.

Features of the AMD Radeon HD 7800 series

The Radeon HD 7800 graphics processor (Pitcairn) has about 2.8 billion transistors and a Graphic Core Next microarchitecture. As mentioned above, the Radeon HD 7850 chip (Pitcairn Pro) has 16 computing units, and its maximum TDP is 130 watts. For the Radeon HD 7870 (Pitcairn XT), these figures are 20 and 175, respectively.

The slide below shows the main specifications of the Radeon HD 7850 and HD 7870 video cards

2 GB of GDDR5 memory is already becoming standard for most mid- and high-end models. Thanks to 256-bit. bus and a high clock frequency of 1200 MHz (4800 MHz effective), the bandwidth is 154 GB/s. This will have a positive effect on performance in games with high resolution and picture quality.

PCI Express 3 interface

In the second half of 2011, almost all motherboard manufacturers introduced their motherboard models with the 3rd generation PCI Express interface. With the release of the Radeon HD 7000 series, video cards with this interface also appeared. PCI Express 3 has twice the bandwidth (32 Gb/s) of the previous generation PCI Express. Compared to PCIe 2, the bandwidth per lane has been doubled from 500 MB/s to 1 GB/s.

Naturally, to take advantage of the new PCIe 3, you need not only a video card and motherboard with this interface, but also support from the processor (not all models from the Ivy Bridge family will support PCIe 3).

Eyefinity 2.0

AMD has gone further in developing its Eyefinity technology, which is designed to display images on multiple monitors. Thanks to the high processing power of the HD 7000 series and support for Eyefinity 2.0, it is now possible to display images on multiple monitors with a total resolution of 16000 x 16000. This allows you to display images on 5 displays with a resolution of 2560x1600 installed in landscape orientation. To work with such resolutions, the older models of the family are equipped with a record 3 GB of GDDR5 (HD 7970 and HD 7950).

AMD Catalyst drivers will support custom resolution starting with the February releases. That is, you can set the required resolution depending on the configuration of the displays in Eyefinity. As of Catalyst 12.2, there is an option to install the Start menu on the display that is most convenient for you, and not on the far left, as it was before. In addition, Eyefinity 2 supports image output in HD3D stereo mode. The combination of three monitors that operate in 3D mode is supported.

Improved tessellation

AMD's Radeon HD 7000 family of graphics cards feature a ninth-generation tessellator and have seen significant performance gains when processing geometry in modern games. The GCN core still includes two Graphics Engines, but where they once contained tessellation and rasterization units, they now consist of an arbitrary number of pipelines dedicated to geometry and pixel processing.

AMD Radeon HD 7800 video cards support the HDMI 1.4a interface, which allows you to output a 120 Hz image (60 Hz per eye), which allows you to display 3D images. With earlier versions of HDMI this was not possible. Starting in December, AMD enabled the ability for HD3D and Eyefinity to work together in drivers.

DirectX 11.1

Radeon 7000 family video cards will support the future DirectX 11.1. It’s too early to say what this will give in practice, since DX 11.1 will be released alongside Windows 8. The main advantages of the new API are outlined as follows:

Independent rasterization;
Flexible combination of graphics computing and video processing;
Native Stereo 3D support.

AMD Unified Video Decoder

It is a hardware part of AMD GPUs responsible for decoding the video stream. UVF has received some improvements in the Radeon 7000 series. In general, UVD retained all the functions of its predecessors, namely support for H.264/AVCHD, MPEG-2, MPEG-4/DivX, VC-1/WMV profile D, Multi-View Codec (MVC), Video Codec Engine ( VCE), AMD Steady Video 2.0. Added support for Dual Stream HD+HD format.