The Radeon HD4870 graphics card is the new king in the top-middle class. Radeon HD4870 graphics card - the new king in the top-middle class Cooling efficiency and overclocking

Introduction The way to increase the performance of the graphics subsystem of the modern PC gaming platform through the use of multi-GPU technologies, which allow combining several graphics processors into a single high-performance complex, has always been considered marginal. In a historical perspective, none of these attempts has become truly successful and has not received widespread acceptance, and this statement can be attributed not only to prehistoric artifacts, such as 3dfx Voodoo2 SLI or ATI Rage MAXX, but also to fairly modern ATI solutions and Nvidia. At some stage, the use of such technologies made it possible to obtain some, sometimes quite substantial, performance gain (often accompanied by a number of technical and software problems), but time passed, and the new generation of single-processor video cards invariably exceeded the achievements of multi-GPU systems built using processors of the previous generations.

However, nothing can last forever. As it turned out relatively recently, the appetites of modern game engines are growing much faster than progress is being made in the field of creating monolithic graphics chips capable of satisfying these appetites. The appearance of the ATI Radeon HD 3870 X2 can be considered the first bell, although in many respects it was due to the lack of competitive graphics chips for the upper market segment in ATI's arsenal at that time. This dual-processor card has shown good results in real conditions and demonstrated that, in addition to the classic, there is an alternative way to create high-performance high-end video adapters. The irony of fate, because it was the absence of a powerful monolithic core at ATI's disposal that helped the company see this path, which may be destined to become the main one in the future. Yes, in the ATI Radeon HD 3870 X2 review, there were words that the bet on multiprocessor graphics systems can only be a tactical move, that it provides less freedom of maneuver, that it is a forced step on the part of ATI, but the verdict on super-powerful monolithic cores has already been voiced.

Nvidia also tried to create a high-end dual-processor solution, but the attempt was not very successful - the GeForce 9800 GX2 did not get rid of all the problems, and the company again focused its main efforts on creating another monolithic monster chip. The latter, in the end, was created, but soon, immediately after the announcement of new ATI solutions, it turned out that, despite the huge size, monstrous complexity and high level of power consumption, the Nvidia G200 not only does not provide a cardinal breakthrough in performance, but also in a number of cases is not superior to the much simpler ATI RV770 or even inferior to it! The verdict went into effect, and Nvidia was forced to sharply cut prices for the GeForce GTX 200 family to at least ensure its competitiveness.

The graphics arm of Advanced Micro Devices, formerly ATI Technologies, has kept Olympic calm with its new strategy. Officially priced at $ 200-300, its solutions provided excellent performance and well-deserved popularity, while the company was preparing to deliver the final blow and finally remove Nvidia from the throne of the king of 3D graphics, preparing the release of the next generation ATI Radeon HD X2. Its announcement was postponed to August 12: despite the fact that the RV770 was released on June 25, there was nowhere to rush, because in any case, Nvidia could not respond with anything to such a step. With such a solid time head start ATI got every chance to bring the new generation of dual-processor cards to perfection, and now it's time to find out how well it used them.


Since the ATI Radeon HD 4870 X2 was originally aimed at the "over $ 500" price segment, in order to avoid a gap between it and the "under $ 400" segment, it was decided to release the less expensive ATI Radeon HD 4850 X2.

Thus, ATI managed to cover all price ranges, from 199 to 549 dollars inclusive, and give Nvidia a battle where the latter remained in the leading positions until recently. To what extent AMD's graphics division managed to win this battle, this review aims to find out. Meet the ATI Radeon HD 4870 X2!

ATI Radeon HD 4870 X2: specifications and innovations

From the point of view of the general concept, the new ATI development, which should confirm its primacy as the developer of the fastest graphics card in the world, differs little from the ATI Radeon HD 3870 X2, except that it uses the new generation RV770 graphics cores instead of RV670 and GDDR5 memory instead of GDDR3. Its technical characteristics in comparison with rivals and younger brother are as follows:

Firstly, twice the amount of video memory is striking than that of the ATI Radeon HD 3870 X2 - it is 2 GB for both two-chip models based on the RV770, although, of course, for technical reasons, applications will only be able to operate with half the volume equal to 1 GB. Here there is a formal parity with the Nvidia GeForce GTX 280, and taking into account the more efficient use of the available local memory by ATI solutions - even a slight advantage. ATI Radeon HD 4800 X2 should fully demonstrate itself in next-generation games and / or at extreme resolutions over 1920x1200, but this statement will be tested in practice below. The total bandwidth of the memory subsystem of the older model exceeds 200 GB / s, but the younger one also boasts a good indicator of 128 GB / s, although this is slightly less than the Nvidia GeForce GTX 280.

Secondly, the computing capabilities of the new products inspire unwitting respect only by the number of ALUs - 1600. According to ATI, the older model of the new family is capable of developing performance over 2.4 teraflops, which is an absolute record in the industry. It should be noted that in double precision calculations, the performance of the ATI Radeon HD 4870 X2 will be about half as much, but this is much more than what the Nvidia GeForce GTX 280 can offer. As for the capabilities of the ATI Radeon HD 4800 X2 in the field of working with textures and rasterization / anti-aliasing - the numbers given in the table indicate parity with the flagship solution from Nvidia, however, in practice, the texture processors of the latter are not so effective, therefore, here too, one should expect higher achievements from new products.

All the aforementioned improvements are not so interesting, as they are, in fact, a simple increase in capacity, in which, of course, there is no revelation. But the ATI Radeon HD 4800 X2 also has something new. The first indication is the type of interface - while the previous generation of ATI dual-processor cards only supported PCI Express 1.1, the current one is fully capable of taking advantage of PCI Express 2.0 thanks to the use of an updated switch chip. This alone expands the "communication" channel between two GPUs twice (from 2.5 to 5 GB / s in each direction), which is useful in modern conditions, when many rendering techniques used in games require data transfer between graphics cores operating in multi-GPU mode.

But the ATI Radeon HD 4800 X2 also has something that the ATI Radeon HD 3870 X2 did not have. This is an additional data transmission channel, the so-called Sideport interface. It links both graphics cores directly and provides an additional 5 GB / s. in each direction, theoretically bringing the total bandwidth of ATI Radeon HD 4800 X2 internal interfaces up to 21.8 GB / s, effectively eliminating all possible bottlenecks and improving the scalability of CrossFireX technology. However, in modern games, the additional communication channel is unlikely to give serious advantages to the ATI Radeon HD 4800 X2, according to ATI's assurances, and this is probably why Sideport is still disabled at the software level, but in the future it may well be in demand.

At the time of the announcement, the new family included two ATI Radeon HD 4800 X2 models, differing in clock speeds and type of memory used, with a recommended price of 549 dollars for the older model 4870 X2 and 399 for the younger model 4850 X2 (the real retail price of the older model in Moscow stores at the time of preparation of the article was about 17 thousand rubles, the youngest has not yet appeared on sale). The senior version, equipped with GDDR5 memory, and of the greatest interest, got into our test laboratory, since it is ATI Radeon HD 4870 X2 that is called upon to take the crown from Nvidia GeForce GTX 280 and fans of ATI products pin all their hopes on it. The ATI Radeon HD 4870 X2 sample was provided by Tul, which supplies graphics cards under the well-known PowerColor brand. Before moving on to the anatomy of the novelty, let's evaluate its packaging and equipment.

PowerColor HD 4870X2 2GB GDDR5: packaging and contents

The PowerColor HD 4870X2 2GB GDDR5 video adapter (model AX4870X2 2GBD5-H, hereinafter simply PowerColor HD 4870X2), despite its belonging to the class of high-end graphics solutions, comes in retail sale in a medium-sized box, so the buyer does not have to pay additional costs supplier for warehousing. The design of the box is nothing extraordinary - another girl (or boy?) With a sword adorns its front side. Frankly, such a drawing is one of the most boring buyers of design stamps, which graphic card manufacturers love to abuse, and therefore almost does not attract attention.

From the point of view of information content, the packaging is not bad, although PowerColor did not fail to mention the presence of 2 GB of video memory, which, as you know, in the case of homogeneous multi-GPU solutions, to which the ATI Radeon HD 4870 X2 belongs, is not true - three-dimensional applications are available only half of the total memory, that is, in this case, 1 GB. This is not a cause for concern, since the flagship solution from Nvidia, the GeForce GTX 280, has the same amount of local video memory, and this is a normal amount for a modern graphics adapter that claims a place in the niche of high-performance high-end cards.

Inside the box there is a cardboard pallet, in the compartments of which the contents are placed - the video adapter itself and related accessories. The latter include the following components:

DVI-I → D-Sub adapter
DVI-I → HDMI adapter
Mini-DIN → YPbPr adapter
Mini-DIN → Composite adapter
CrossFire connecting bridge
Quick Installation Guide
Driver disc

The package bundle is strange not only for a product costing more than 15 thousand rubles, but also for 2008 as a whole - for example, it is unclear why the kit contains two whole adapters for connecting in analog YPbPr formats and, especially, Composite. If the first still provides acceptable image quality, then the second is hopelessly outdated and is practically not used in modern TVs, with the exception of very cheap models. Alas, there is also no power adapter for an 8-pin connector - although it is physically compatible with 6-pin power supply connectors, the card will not work with this connection. Many power supplies, even very powerful models released in recent months, do not have 8-pin connectors.

As, alas, it often happens, there is no software for playing HD video that supports the advanced capabilities of the ATI UVD 2 video processor. Of course, such a player can retail for $ 50 and more, but the OEM version, intended for graphic card manufacturers, cost would be much cheaper and unlikely to have a significant impact on the retail price of the final product. Moreover, for a modern card with a similar price, we consider the presence of such software in the bundle necessary - a buyer who spends such a significant amount has the right to be sure that he will be able to use all the features of the purchased product without additional financial costs.

In general, the packaging of PowerColor HD 4870X2 does not cause any complaints: although it cannot boast of an original design, it has acceptable dimensions and can easily fit into a regular plastic bag. But the complete set, taking into account the high cost of this solution, could be richer. In any case, the presence of a software player that supports HD video and hardware capabilities of ATI graphics cores for its decoding and processing would not be superfluous.

ATI Radeon HD 4870 X2: PCB Design

The ATI Radeon HD 4870 X2 is an extremely complex design that requires two GPUs, associated memory kits, power subsystems, and a PCI Express switch on a single PCB. It is extremely difficult to design such a graphics adapter, but the ATI development team has a background in the creation of the ATI Radeon HD 3870 X2, and they did an excellent job of the task. Despite the fact that the novelty carries two RV770 cores on board, it does not exceed the dimensions of the single-processor Nvidia GeForce GTX 280 - the length of both printed circuit boards is 27 centimeters:

Of course, 27 centimeters is a lot, and the novelty can not be installed in any case, but since it belongs to the most productive class of graphics cards, hardly anyone will try to install it in a microATX system. At the same time, the ATI Radeon HD 4870 X2 does not look as massive as the Nvidia GeForce GTX 280, because, unlike the latter, it is not enclosed in a blank metal casing. It is interesting to note that, unlike other ATI solutions, the dual-processor flagship uses a black board instead of red. Most likely, this was done in order to make it as impressive as possible - a kind of hint that ATI Radeon HD 4870 X2 belongs to the elite of 3D graphics.

Unlike the Nvidia GeForce GTX 280, the ATI Radeon HD 4870 X2 cooling system is easy to dismantle, giving access to the "insides" of the new product, which is of the greatest interest from a technical point of view. The PCB layout as a whole is very similar to the ATI Radeon HD 3870 X2, which is not surprising given the close relationship of these graphics cards. Actually, the novelty is an evolutionary development of the latter.

Although the back of the board is still empty, the ATI Radeon HD 4870 X2's power system has been significantly redesigned and strengthened, which is logical given the increased power consumption of the RV770 compared to the RV670. It includes two three-phase regulators with the familiar Volterra VT1165MF controllers, often used in high-frequency graphics card power converters. Each set of memory chips is powered by a separate regulator; unlike the ancestor, there is no room for further increasing the capacity of the power supply system.

It is known that the peak power consumption of a single ATI Radeon HD 4870 can reach 130 W, respectively, for an ATI Radeon HD 4870 X2, in fact, combining two such cards in one printed circuit board, this figure should be multiplied by at least two, and 260 W is a very serious number. In this light, the presence of one eight-pin PCI Express 2.0 power connector on the board looks more than justified. Most likely, it will carry the main load, which is evidenced by a simple fact: unlike its ancestor, ATI Radeon HD 4870 X2 requires the obligatory connection of an eight-pin power supply cable to the corresponding connector, otherwise the card does not start, displaying a message about the presence of problems with meals.

As practice has shown, the card can be made to work with two six-pin power connectors - for this, it is enough to close the remaining free “legs” of the 8-pin connector to ground (in fact, the adapter from 6-to 8-pin connector does the same ), but of course, the stability of such work is not guaranteed and depends on the quality of both specific connectors and the power supply unit as a whole. The location of the connectors also raises some complaints - they are directed downward, which makes it difficult to connect the power cables to the card already installed in the system.

The function of managing the exchange of data between two RV770 cores is assigned to the PLX Technology PEX8647 chip. This chip is a 3-port PCI Express Gen 2 switch that supports PCIe 2.0, which means twice the data transfer rate. Like its predecessor, the PEX8547, the new switch has direct data exchange (so-called peer-to-peer mode), which allows graphics cores to communicate directly, bypassing the PCI Express root controller included in the system chipset. Despite the improved specifications, the switch has an extremely low power consumption of only 3.8W, while its ancestor consumed about 5W at lower data rates; moreover, the new chip also has significantly smaller dimensions.

In addition to using the capabilities of the switch, GPUs can communicate directly using the ATI CrossFireX interface or an additional data transfer channel, the so-called Sideport, operating at the same speed - 5 GB / s in each direction, however, at the moment Sideport support is disabled by software. Since the ATI Radeon HD 4870 X2 is, in fact, a bundle of two ATI Radeon HD 4870s, the cores are also linked by the traditional CrossFire channel, so the board has only one external connector for this interface to support four-processor configurations. More than two ATI Radeon HD 4870 X2 as part of a single complex cannot be used for purely technical reasons, and even if it were possible, the level of power consumption of such a graphics system would be truly monstrous.

The RV770 graphics cores are located to the right and left of the switch, and, unlike the ATI Radeon HD 3870 X2, the developer company decided not to raise their frequency, apparently for reasons related to the level of power consumption. As a result, the frequency of the chips is the same as in a single ATI Radeon HD 4870, and amounts to 750 MHz. The core configuration is standard: 800 ALUs grouped into 160 superscalar computational units, 40 texture processors, and 16 RBEs. Thus, the total configuration of the ATI Radeon HD 4870 X2 is as follows: 1600 ALU, 80 texture processors and 32 RBE - theoretically, this is more than enough to put any existing Nvidia video adapter on both blades, including the GeForce GTX 280 and GeForce 9800 GX2. Unlike the ATI Radeon HD 3870 X2, the ATI Radeon HD 4870 X2 has no bottlenecks, at least at first glance. The only such place may be insufficiently thorough software support for CrossFire in drivers, but this is typical of any modern multi-GPU system, be it developed by ATI or Nvidia (just remember the results of testing a Quad SLI system). In the worst case, only one graphics chip will work, and the RV770, even on its own, is able to form a very serious competitor to solutions based on the Nvidia G200.

The titanic computing power of the ATI Radeon HD 4870 X2 is only half the contribution to its future gaming victory. The other half is the memory subsystem, a component that has a significant impact on high-resolution performance. From this "board" the novelty is armed no worse, since it carries two sets of GDDR5 chips (2x8 chips) with a total volume of 2 GB - 1 GB for each graphics core. Unlike the usual ATI Radeon HD 4870, which is equipped with Qimonda IDGV51-05A1F1C-40X memory, the dual-core ATI Radeon HD 4870 X2 is equipped with Hynix H5GQ1H24MJR-T0C chips with a capacity of 1 Gbit (32Mx32). They are designed for a supply voltage of 1.5 V and are capable of operating at a frequency of 1000 (4000) MHz, but their actual operating frequency is 900 (3600) MHz. Since we are talking about a homogeneous type of dual-GPU graphics system, the total amount of video memory available for 3D applications is 1 GB. Thus, the ATI Radeon HD 4870 X2 should not suffer from a lack of video memory, especially since ATI solutions use the available local memory more efficiently than Nvidia solutions. The total bandwidth of the memory subsystem of the new item is 230.4 GB / s, much more than what Nvidia can offer today - at least when it comes to single graphics cards. Indeed, ATI's main rival should consider its new solution from the position of "armed and very dangerous", since, as already mentioned, it is practically devoid of weak points.

The connector configuration meets the established standard and includes two dual-link DVI-I ports with support for resolutions up to 2560x1600, and, unlike Nvidia SLI solutions, multi-monitor configurations are fully supported. In addition to them, there is also a universal seven-pin mini-DIN connector, which in our time of HDMI dominance is practically not in demand. As mentioned above, the card is equipped with one CrossFire connector, which allows it to be used as part of a four-processor graphics system, combining it with another ATI Radeon HD 4870 X2.

ATI Radeon HD 4870 X2: cooling system design

Even removing 170 W of heat is a rather serious task, especially taking into account all the restrictions imposed on the design of the graphics adapter cooling system, and the ATI Radeon HD 4870 X2 should have a significantly higher heat dissipation level than its predecessor, ATI Radeon HD 3870 X2. In this light, the design of its cooling system is of particular interest.

Oddly enough, ATI did not consider the problem significant and did not seriously change the design of the ATI Radeon HD 4870 X2 cooling system. It is still based on two solid-milled radiators, although, unlike the ATI Radeon HD 3870 X2, both of them are copper. Such radiators are quite effective, since their ribs form a single whole with the base, and, unlike composite radiators, there is no additional thermal resistance at the point where the ribs are attached to the base - but it is still very unusual to see the actual absence of heat pipes (except for a full-fledged tube heat chamber on one of the radiators) in a product with a predicted heat dissipation level of 260 W. The heatsinks contact the GPU crystals through a layer of classic dark gray thermal paste.

With this design of the cooling system, ATI is quite at risk, even if it decided to sacrifice a comfortable noise level in favor of improving cooling efficiency. The heatsinks are attached with four screws and a spring-loaded Phillips plate on the back of the PCB. Although they are not rigidly connected to the base, they still have fairly tight contact with it, and, it seems, can transfer part of the heat flux to it. In theory, this should slightly increase the overall cooling efficiency.

The base itself acts as a heat sink for the power elements of the power supply system, memory chips located on the front side of the printed circuit board, and a PCI Express switch chip. It also serves as a mounting point for the fan. To ensure reliable thermal contact, elastic rubber-like gaskets are used; their efficiency is low, but the heat dissipation levels of the aforementioned elements are not as high as in graphics cores. The top of the base has a series of needle-like protrusions that increase its efficiency as a radiator. The base is attached to the board separately by means of 8 screws. The memory chips located on the back of the PCB are cooled by a separate aluminum plate.

The fan is the same as in the ATI Radeon HD 4870 - NTK Technologies CF1275-B30H-C004 cooling system. This radial turbine has a maximum current of 1 A, that is, with a supply voltage of 12 V, it can consume up to 12 W. With such a power, it will surely cope with blowing out even two radiators, the only question is, at what level of noise will this be achieved. The release of hot air is carried out through the slots in the mounting plate of the video adapter, and with an energy consumption level of about 260 W, it will be able to claim the role of an additional source of heat in the room, which can be useful in winter, but will hardly please anyone in the summer. The profiling casing is made of dark brown smoky translucent plastic, and in the area of ​​the needle radiator is equipped with a sticker warning about the high temperature of the latter.

Overall, the ATI Radeon HD 4870 X2 cooling system is surprisingly simple for a card with a heat dissipation level around 260 W. It remains to be hoped that it will cope with its task, at least at the cost of some deterioration in noise characteristics.

ATI Radeon HD 4870 X2: Power Consumption, Thermal, Noise and Compatibility

The information on the power consumption level of the ATI Radeon HD 4870 X2 is of great interest, if only because this is the first video adapter, whose power consumption level must surely exceed 200 watts. As our readers know, until now, none of the single graphics cards have been able to overcome this barrier, despite the statements of the development companies. To carry out the necessary measurements, a specially equipped test bench was used with the following configuration:

AMD Athlon 64 FX-55 processor (2.6 GHz)
EPoX EP9-NPA + SLI Motherboard (Nvidia nForce4 SLI)
PC3200 memory (2x512 GB, 200 MHz)
Western Digital Raptor WD360ADFD Hard Drive (36GB, SATA-150, 16MB Buffer)
Chieftec ATX-410-212 power supply (rated power 410 W)

Futuremark PCMark05 Build 1.2.0
Futuremark 3DMark06 Build 1.1.0

As usual, the first SM3.0 / HDR test of the 3DMark06 package was used to create a load in 3D mode, running in a loop at 1600x1200 with forced MSAA 4x and AF 16x. The test is excellent for this purpose, and, as practice shows, it loads the GPU even better than the tests of the more modern 3DMark Vantage, so there is no particular reason to refuse to use it yet. The "peak 2D" mode was emulated using the 2D Transparent Windows test included in PCMark05. This test is quite relevant, since it emulates active work with windows, and the Windows Vista Aero window system, as you know, uses 3D functions. It is possible that in the future this test will be replaced by a test of playing HD-video, as a more urgent task for an ordinary user.

As a result of the measurements, the following data were obtained:

The peak power consumption of the ATI Radeon HD 4870 X2 has indeed reached the predicted value, thus winning the title of the most uneconomical graphics card. The figures in 2D and Peak 2D modes, of course, are quite high, but they are not too different from those of the usual ATI Radeon HD 4870, but the result in 3D is simply monstrous and surpasses everything we have seen before. This card will really require a powerful power supply, and the presence of a "native" eight-pin PCI Express 2.0 connector on the unit is highly desirable, since it carries the main load, which almost reaches the maximum value of 150 watts.

Of course, such a high level of power consumption has a corresponding effect on the heat dissipation level of the ATI Radeon HD 4870 X2. Unfortunately, RivaTuner does not yet support the new product, but even according to the Catalyst Control Center, the card operates in a rather intense thermal mode, the temperature of the GPUs fluctuates from 64 ° C in idle to 86-90 ° C during active work in 3D applications. At first glance, the numbers are not high, but it is not known what exactly the corresponding CCC panel displays, since the board has two GPUs, and both of them contribute equally to the level of heat dissipation. We only know that in idle mode, the clock speed of the master core drops to 500 MHz, and the slave core is probably turned off completely or almost completely, which is confirmed by the results of measurements of the power consumption level. The high temperature warning symbol flaunts on the casing of the cooling system for a reason - subjectively, the card heats up very much, and after any length of time in 3D mode it is almost impossible to pick it up, since the temperature of the metal parts clearly exceeds the pain threshold.

Noise level measurements carried out with the Velleman DVM1326 sound level meter showed the following picture:

The results are not very encouraging: although the card makes little noise in 2D mode, under load the fan ramps up in order to cope with the blowing of two heatsinks generating a total heat flow of 260 W, and the noise level increases significantly, almost reaching the values ​​previously demonstrated by the ATI cooling system Radeon HD 2900 XT. Subjectively, the spectral composition of the noise is quite favorable, the main contribution to it is made by the noise of the air flow passing through the radiators, and not the sounds emitted by the turbine, but, unfortunately, the card suffers from the same drawback as a single ATI Radeon HD 4870. Control system fan turns excessively "intellectually", periodically changing its speed, and the noise level fluctuations caused by this are very annoying to the ear.

Despite the use of a PCI Express switch chip, the ATI Radeon HD 4870 X2 has no problems with compatibility with various motherboards and chipsets - both modern ones with PCI Express 2.0 support, and older ones that only support PCI Express 1.0a or 1.1.

Test platform configuration and test methodology

To study the practical potential of ATI Radeon HD 4870 X2, we used test platforms with the following configuration:

Intel Core 2 Extreme X6800 processor (3.0 GHz, FSB 333 MHz x 9)
DFI LANParty UT ICFX3200-T2R / G (ATI CrossFire Xpress 3200) Motherboard for ATI Radeon HD
Asus P5N32-E SLI (Nvidia nForce 680i SLI) motherboard for Nvidia GeForce
Corsair TWIN2X2048-8500C5 memory (2x1GB, 1066MHz, 5-5-5-15, 2T)
Maxtor MaXLine III 7B250S0 Hard Drive (250GB, SATA-150, 16MB buffer)
Power supply unit Enermax Galaxy DXX EGX1000EWL (maximum power 1000 W)
Dell 3007WFP Monitor (30 "Max Resolution [email protected] Hz)
Microsoft Windows Vista Ultimate 32-bit
ATI Catalyst 8.7 for ATI Radeon HD 4800
ATI Catalyst 8.52.2.0 Sample for ATI Radeon HD 4870 X2
Nvidia GeForce 177.41 WHQL for Nvidia GeForce

The drivers have been tuned according to a standard procedure to provide the highest possible texture filtering quality while minimizing the impact of software default optimizations. Anti-aliasing for transparent textures has also been enabled. As a result, the settings for ATI and Nvidia drivers began to look like this:

ATI Catalyst:

Catalyst A.I .: Standard
Mipmap Detail Level: High Quality
High Quality AF: On
Wait for vertical refresh: Always Off
Enable Adaptive Anti-Aliasing: On / Quality
Method: Multi-sampling
Temporal Anti-Aliasing: Off

Nvidia GeForce:

Texture filtering - Quality: High quality
Texture filtering - Trilinear optimization: Off
Texture filtering - Anisotropic sample optimization: Off
Vertical sync: Force off
Antialiasing - Gamma correction: On
Antialiasing - Transparency: Multisampling
Other settings: default

The test package included the following set of games and applications:

3D first-person shooters:

Battlefield 2142
BioShock
Call of juarez
Call of duty 4
Crysis
Enemy Territory: Quake Wars
Half-Life 2: Episode Two
S.T.A.L.K.E.R .: Shadow of Chernobyl

3D shooters with a third person view:

Lost Planet: Extreme Condition
Tomb Raider: Legend

RPG:

Hellgate: London
The Elder Scrolls IV: Oblivion

Strategies:

Company of Heroes: Opposing Fronts
Command & Conquer 3: Tiberium Wars
World in Conflict

Synthetic tests:

Futuremark 3DMark06
Futuremark 3DMark Vantage

Each of the games included in the set of test software was tuned to the highest possible level of image detail by using tools available in the game itself to any uninitiated user. This means a fundamental rejection of manual modification of any configuration files, since the user is not required to be able to do this. The exception is the shooter Enemy Territory: Quake Wars, which had the built-in average performance limiter disabled, fixed at around 30 frames per second. Games that can use the capabilities of DirectX 10 were tested in this mode.

In addition to the ATI Radeon HD 4870 X2, the following single graphics cards took part in the tests:

ATI Radeon HD 4870 (RV770, 750/750/3600 MHz, 800 SP, 40 TMU, 16 RBE, 256-bit 512 MB GDDR5)
ATI Radeon HD 4850 (RV770, 625/625/2000 MHz, 800 SP, 40 TMU, 16 RBE, 256-bit 512 MB GDDR3)
Nvidia GeForce GTX 280 (G200, 600/1300/2200 MHz, 240 SP, 80 TMU, 32 RBE, 512-bit 1024 MB GDDR3)
Nvidia GeForce GTX 260 (G200, 576/1242/2000 MHz, 192 SP, 64 TMU, 28 RBE, 448-bit 896 MB GDDR3)
Nvidia GeForce 9800 GX2 (2xG92, 600/1500/2000 MHz, 256 SP, 128 TMU, 32 RBE, 2x256-bit 2x512 MB GDDR3)

Additionally, ATI Radeon HD 4870 and ATI Radeon HD 4850 have been tested in CrossFire configuration.

Since ATI Radeon HD 4870 X2 belongs to the highest price category, the standard list of modes has been supplemented with 2560x1600 and 2048x1536 resolutions; the latter is intended for games that do not support 16:10 aspect ratio. In almost all cases where the use of MSAA 4x is possible without disabling the SM3.0 / 4.0 / HDR effects, anti-aliasing supplemented the standard 16x anisotropic filtering, which is used by default in all tests except 3DMark. Anti-aliasing and anisotropic filtering were activated either by the game itself, or, if they were absent, it was forced using the appropriate settings of the ATI and Nvidia drivers.

To obtain performance data, either the tools built into the game were used, with the obligatory recording of the original demos, if possible, or, if they were not available, the Fraps 2.9.1 utility in manual mode. Whenever possible, data were recorded not only on the average, but also on the minimum performance.

Game tests: Battlefield 2142

The game does not officially support 16:10 aspect ratio, so it uses resolutions with the classic aspect ratio of 4: 3.

Such a powerful video adapter, which is the ATI Radeon HD 4870 X2, has nowhere to prove itself, but even despite the disabled Sideport interface, it demonstrates a slightly higher minimum performance compared to the ATI Radeon HD 4870 CrossFire tandem. This is probably the merit of the PCI Express switch, which allows GPUs to communicate directly, bypassing the bus root controller, and therefore avoiding additional delays in data transfer.

Game tests: BioShock

BioShock does not support FSAA when running on Windows Vista in DirectX 10 mode, so the game is tested without anti-aliasing.

Despite the fact that the game runs on the relatively undemanding engine Unreal Engine 3 and does not use full-screen anti-aliasing, the ATI Radeon HD 4870 X2 performs excellently, demonstrating over 100 frames / s at 2560x1600 and almost 75% ahead of Nvidia's fastest solution, GeForce GTX 280. At the same time, unlike the classic pair of ATI Radeon HD 4870, the new product is completely stable in operation and does not experience any problems with the CrossFire technology.

Game tests: Call of Juarez

The game does not support resolutions higher than 1920x1200, so we couldn't get results for 2560x1600.

The average result of ATI Radeon HD 4870 X2 unexpectedly turned out to be slightly less than the similar result of two separate ATI Radeon HD 4870s operating in CrossFire mode, but the minimum performance remained practically unchanged, and nothing prevents you from using the card in resolutions up to 1600x1200 / 1680x1050 without sacrificing gaming comfort. Taking into account the fact that Call of Juarez actively uses modern rendering techniques, it makes sense to expect a noticeable performance increase in the ATI Radeon HD 4870 X2 after the software blocking of the Sideport interface is removed.

Game tests: Call of Duty 4: Modern Warfare

The main advantage of ATI Radeon HD 4870 X2 in this case is a higher minimum performance compared to ATI Radeon HD 4870 CrossFire, especially at 2560x1600 resolution, where, apparently, twice the amount of local video memory has a beneficial effect on the result. ATI's solution loses out to Nvidia GeForce GTX 280 by only 6%, which does not affect the comfort level in any way - both cards can be successfully used at 2560x1600 with MSAA 4x enabled.

Game tests: Crysis

The game is tested at High detail settings, with the exception of the Shaders option set to Very High. This is a trade-off between image quality and performance.

Alas, the expected breakthrough in performance has never happened, and the inclusion of Sideport is unlikely to make it possible. Nevertheless, the results of ATI Radeon HD 4870 X2 are not bad at all, and it is noticeably ahead of Nvidia GeForce GTX 280. The gap is especially large at 2560x1600, where the advantage of ATI Radeon HD 4870 X2 in average performance reaches an impressive 45%. ATI's account should also be attributed to excellent stability, which the classic CrossFire tandem cannot boast of.

Game Tests: Enemy Territory: Quake Wars

Since the official performance in the game is fixed at 30 frames / s, and the update of the physical model on the server also occurs at a frequency of 30 Hz, this figure is the minimum acceptable for Quake Wars.

In normal resolutions, ATI Radeon HD 4870 X2 is barely noticeably inferior to the ATI Radeon HD 4870 CrossFire bundle, but on the whole, their results can be called equal, without erring against the truth. However, in 2560x1600 resolution, the new ATI development decisively breaks ahead, leaving behind all rivals and confirming its title of the new king of 3D graphics. The reason is obvious - the MegaTexture technology used in this game operates with textures with a resolution of 32768x32768 and an uncompressed volume of about 3 GB, so the presence of 1 GB of local video memory is very appropriate here.

Game tests: Half-Life 2: Episode Two

Combining two RV770 chips on one board by means of a PCI Express switch gives little in Episode Two, and only at 2560x1600 the advantage of a single-board dual-chip card over the classic tandem of two single-chip cards reaches 5%.

At the same time, the ATI Radeon HD 4870 X2 continues to prove its reputation as a champion, since the Nvidia GeForce GTX 280 demonstrates a noticeably worse level of average performance in the same resolution. However, do not forget about the monstrous level of power consumption of the novelty, exceeding 260 W - such is the price to pay for the first place.

Game tests: S.T.A.L.K.E.R .: Shadow of Chernobyl

Since when using a dynamic lighting model, the game does not support FSAA, and when the static model is turned on, it greatly loses its visual appeal, S.T.A.L.K.E.R. tested with anisotropic filtering only.

The ATI Radeon HD 4870 X2 does not demonstrate anything new in S.T.A.L.K.E.R. either, since 1 GB of video memory is not required here, and other technical characteristics of the new item coincide with the characteristics of the usual ATI Radeon HD 4870 CrossFire tandem. Nevertheless, the first place belongs to him, and at 2560x1600 both average and minimum performance are at a quite comfortable level.

Game Tests: Lost Planet: Extreme Condition

There are no significant differences in performance between ATI Radeon HD 4870 X2 and two ATI Radeon HD 4870s operating in CrossFire mode, with the exception of 2560x1600, where the former is 21-22% faster. However, this victory has no practical significance, since the average performance at 17 frames / s cannot be called comfortable under any circumstances. Since the result practically does not differ from the result of a single ATI Radeon HD 4870, the inevitable conclusion follows - either the CrossFire technology support in this game does not work correctly, despite some increase in resolutions below 2560x1600, or its full-fledged work is impossible in principle due to the peculiarities of the game engine ...

Game tests: Tomb Raider: Legend

With the advent of the ATI Radeon HD 4870 X2, the primacy of Nvidia solutions was shaken, although not in all resolutions - the new product outstripped the GeForce GTX 280 for some reason only in 1600x1200 and 2560x1600, slightly yielding to the latter in 1280x1024 and 1920x1200. Also, unlike the usual ATI Radeon HD 4870 CrossFire tandem, there were no problems with the operation of the multi-GPU mode; in the case of CrossFire, we never managed to activate it - despite all attempts, there was no performance gain.

Game Tests: Hellgate: London

The game is distinguished by increased appetites for video memory, which is confirmed by the results of Nvidia GeForce GTX 280, but ATI Radeon HD 4870 X2 has a similar amount of available local memory, and it manages to get ahead at 2560x1600, although at resolutions below 1920x1200 it lacks stars from the sky , significantly inferior to the development of Nvidia. This is not a disadvantage, since in all cases the performance of the novelty remains at a comfortable level. We can say that the ATI Radeon HD 4870 X2 fulfills its cost.

Game Tests: The Elder Scrolls IV: Oblivion

Without HDR, the game significantly loses its attractiveness, and although the opinions of the players differ on this, the methodology we use implies testing TES IV in the mode with FP HDR enabled.

The ATI Radeon HD 4870 X2 does not make a revolution, but in this case, there is no room for it - the performance ceiling in Oblivion has long been reached, and the game can no longer be considered "heavy" for modern graphics cards. Alas, in open areas at 2560x1600, the minimum performance of ATI's solution may drop slightly below the acceptable level, while this does not happen with the Nvidia GeForce GTX 280. It is possible that activating Sideport will correct this shortcoming, but the chances of this are few.

Game Tests: Company of Heroes: Opposing Fronts

Company of Heroes is tested only in DirectX 10 mode, as it achieves the maximum level of detail, and therefore, the visual experience of the game is the most complete.

Another confident victory for ATI - its development was able to set a record in 2560x1600 resolution, and not only in average, but also in minimum performance. Although the latter never reached 25 fps, the ATI Radeon HD 4870 X2 still rightfully receives the title of the card that provides the greatest degree of comfort in CoH: Opposing Fronts. This is not the limit, and the potential reserve provided by the additional data transfer channel between the RV770 cores should not be discounted.

Game tests: Command & Conquer 3: Kane's Wrath

The add-on to C&C 3: Tiberium Wars brought no technical changes. The game still contains a limiter for average performance, so when comparing graphics cards, their minimum performance should be considered first.

Testing modern high-end graphics cards in Command & Conquer 3 games is no longer of significant interest, since they all show the same result. Soon this position in the list of test software will be replaced by something more relevant.

Game tests: World in Conflict

As in the case of CoH: Opposing Fronts, the ATI Radeon HD 4870 X2 failed to provide a radical breakthrough in performance, but rightfully took first place at 2560x1600, becoming the solution that provides the highest possible level of comfort with MSAA 4x enabled. None of the single Nvidia-designed cards came close to the values ​​demonstrated by the ATI Radeon HD 4870 X2.

Synthetic benchmarks: Futuremark 3DMark06

ATI Radeon HD 4870 X2 did not become the record holder in 3DMark06, but this test suite has not presented any difficulty for modern graphics cards for a long time; moreover, there is nowhere for a solution with such impressive potential as the new ATI development has. In all summary results, it is slightly inferior to the classic ATI Radeon HD 4870 CrossFire tandem, but their results are not indicative due to the use of a resolution of 1280x1024 without anti-aliasing.

The results of individual tests of the SM2.0 group repeat what was seen earlier - the ATI Radeon HD 4870 X2 is somewhat slower than the ATI Radeon HD 4870 CrossFire, but the difference is small and completely uncritical.

In the SM3.0 / HDR test group, a similar situation is observed only in the second test, while in the first test the results of ATI Radeon HD 4870 X2 and ATI Radeon HD 4870 CrossFire coincide with an accuracy of 0.3 fps.

Synthetic benchmarks: Futuremark 3DMark Vantage

To minimize CPU impact, 3DMark Vantage uses an “Extreme” profile that uses 1920x1200 resolution, 4x FSAA and anisotropic filtering.

Unlike 3DMark06, 3DMark Vantage not only actively uses the capabilities of modern graphics cards, it uses all resources, including the amount of video memory, to the maximum. According to the last parameter, ATI Radeon HD 4870 X2 has a twofold superiority over the ATI Radeon HD 4870 CrossFire system, as a result of which its final result is almost 900 points higher. The gap with the Nvidia GeForce GTX 280 is monstrous, and fully confirms the main strategic mistake of Nvidia, when designing a new GPU, it relied on the performance of texture processors with rather modest computing capabilities.

The performance of the ATI Radeon HD 4870 X2 in the first test is 20% higher than the usual CrossFire tandem consisting of two ATI Radeon HD 4870s, even with Sideport disabled. Apparently, the main contribution here comes from the larger video memory, because the Extreme profile uses a resolution of 1920x1200 with MSAA 4x. In the second test, the advantage is less, and is only 12%, but this does not detract from the victory of ATI.

Conclusion

The ATI Radeon HD 4870 X2 testing is over, and it's time to take stock. The results demonstrated by the novelty give rise to the most optimistic, with rare exceptions, estimates. Pivot diagrams will, as usual, help in making the final verdict:

From the very beginning, the Radeon HD 4870 X2 took a fairly high start, snatching victory in 8 tests out of 16, moreover, in Call of Juarez, the advantage over Nvidia GeForce GTX 280 was over 80%, and in two more games it exceeded 15%.

In seven tests, the new product lost, but only in three, including World in Conflict, the lag was significant, and in both CoH: Opposing Fronts and Hellgate: London, the ATI Radeon HD 4870 X2 card provided a comfortable level of performance without any problems. In general, it is difficult to imagine that someone would use such an expensive and powerful card as the Radeon HD 4870 X2 at 1280x1024, so the results shown by him in this mode are not decisive.

Already at 1600x1200 the number of games in which the ATI Radeon HD 4870 X2 loses to the Nvidia GeForce GTX 280 is reduced to only three, and even in these cases its performance remains acceptable for practical use. In all other games, excluding Command & Conquer, ATI's flagship outperformed the previous leader among single cards, and in the minimum case the advantage was only 1.7%, and in the maximum it almost reached 100%.

Approximately the same picture was observed at 1920x1200, but in general the advantage of ATI Radeon HD 4870 X2 over Nvidia GeForce GTX 280 became more confident - the number of cases where it was more than 10-15% increased, and in Call of Juarez the gap even exceeded 100%, that is, more than doubled. To this we should add a convincing victory in 3DMark Vantage, since the Extreme profile used for testing implies exactly the resolution of 1920x1200.

The 2560x1600 resolution became the finest hour of the ATI Radeon HD 4870 X2. The new product was carefully prepared for this, and it did not disappoint - the doubled video memory allowed it to win in all tests, with the exception of Call of Duty 4, but even in it the gap from Nvidia GeForce GTX 280 did not exceed 6%.

Moreover, in almost all games, with the exception of Crysis, Lost Planet and World in Conflict, the Radeon HD 4870 X2 was able to maintain a comfortable or close to that level of performance, which makes it the first graphics adapter that makes 2560x1600 a truly practical use!

For the sake of fairness, it should be noted that not everything is so cloudless and the novelty is not devoid of certain drawbacks. First of all, these include the prohibitively high level of energy consumption, which for the first time in our practice exceeded the 200 W mark, and exceeded it quite significantly. This automatically entailed an equally high level of heat generation, to combat which a comfortable noise level had to be sacrificed. As a result, the payback for the record high level of performance was the need to use a really powerful power supply unit (for a system with one card - at least 500 W, subject to quality workmanship and good voltage stability), the possibility of using an ATI Radeon HD 4870 X2 to heat a room, and also fairly mediocre noise performance.

The classic drawback of homogeneous multiprocessor graphics systems has not disappeared anywhere - the lack of a significant increase in speed with insufficient quality software support. Although we must give ATI its due, this problem has become much less frequent; Based on the data we received, we can even talk about its eradication, but do not forget that there are a lot more games on the market than are included in our test suite, and in some of them the eternal disease of modern multi-GPU technologies can re-manifest itself. We can only hope for ATI's operational work in the field of cooperation with game developers.

On the whole, ATI's claim to be the leader in consumer 3D graphics should be considered completely legitimate - today the ATI Radeon HD 4870 X2 is truly a performance champion. Do not forget that it is not a direct competitor to the Nvidia GeForce GTX 280, and you will have to pay a slightly larger sum for the possession of such a monster. As for the PowerColor HD 4870X2 2GB GDDR5, it is a typical representative of the cohort of reference cards, which, moreover, cannot boast of an extraordinary package. However, by purchasing this product, for your money you will get what you paid for, namely the fastest gaming video adapter for today.

PowerColor HD 4870X2 2GB GDDR5: pros and cons

Advantages:

Best gaming performance to date
Outstanding performance at high resolutions
Outperforms Nvidia GeForce GTX280 in most benchmarks
Wide range of FSAA modes
Industry-leading edge detect CFAA anti-aliasing
Excellent quality anisotropic filtering
DirectX 10.1 and Shader Model 4.1 support
Full hardware support for HD video decoding
High quality HD video post-processing including upscaling
Integrated 8-channel audio engine with support for HD audio formats
Supports HDMI audio output
No compatibility issue

Disadvantages:

Cooling system efficiency in question
High noise in 3D mode
Extremely high power consumption and heat dissipation
Maximum performance depends on software CrossFire support
Poor equipment

Check availability and cost of ATI Radeon HD 4870 X2

Other materials on this topic

ATI Radeon HD 4870 vs. Nvidia GeForce GTX 260 at 2560x1600: Perfection Made More Accessible?
Review of ATI Radeon HD 4850/4870 gaming performance: from theory to practice
ATI Radeon HD 4800 Graphics Architecture Review: Long-awaited Revenge?

HIS RADEON HD 4870 512MB PCI-E

Connection to analog monitors with d-Sub (VGA) is made through special adapters-adapters DVI-to-d-Sub. DVI-to-HDMI adapters are also supplied (we remember that these accelerators support full-fledged video and sound transmission to an HDMI receiver), so there should be no problems with such monitors either.

Maximum resolutions and frequencies:

• 240 Hz Max Refresh Rate
• 2048 × 1536 × 32bit x85Hz Max - analog interface
• 2560 × 1600 @ 60Hz Max - digital interface (all DVI jacks with Dual-Link)

As for the capabilities of video cards to play MPEG2 (DVD-Video), back in 2002 we studied this issue, since then little has changed. Depending on the movie, the CPU load during playback on modern video cards does not rise above 25%.

About HDTV. One of the studies has also been carried out and can be viewed.

Unfortunately, at the moment the RivaTuner utility (by A. Nikolaychuk AKA Unwinder) does not support the new series, and therefore there is no monitoring.

Equipment.

The basic package should include: user manual, CD with drivers and utilities, DVI-to-VGA adapter, CrossFire bridge, DVI-to-HMDI adapter, component output adapter (TV-out), and external power splitters ... Next, we will show what is offered in addition to the map.

Packaging.

Installation and drivers

Test bench configuration:

• Intel Core2 based computer (775 Socket)
  • Intel Core2 Extreme QX9650 processor (3000 MHz);
  • Zotac 790i Ultra motherboard based on Nvidia nForce 790i Ultra chipset;
  • RAM 2 GB DDR3 SDRAM Corsair 2000MHz (CAS (tCL) = 5; RAS to CAS delay (tRCD) = 5; Row Precharge (tRP) = 5; tRAS = 15);
  • hard drive WD Caviar SE WD1600JD 160GB SATA.
  • Tagan TG900-BZ 900W power supply unit.
• operating system Windows Vista 32bit SP1; DirectX 10.1;
• Dell 3007WFP monitor (30 ").
• ATI drivers CATALYST 8.6; Nvidia versions 175.16 (9xxx series) and 177.34 (GTX 2xx).

VSync is disabled.

Synthetic tests

The synthetic test packages we use can be downloaded here:

• D3D RightMark Beta 4 (1050) with a description on the website 3d.rightmark.org
• D3D RightMark Pixel Shading 2 and D3D RightMark Pixel Shading 3- tests of pixel shaders versions 2.0 and 3.0 link.
• RightMark3D 2.0 with a short description:,

RightMark3D 2.0 requires an installed MS Visual Studio 2005 runtime package, as well as the latest DirectX runtime update.

Synthetic tests were carried out on the following video cards:

• RADEON HD 4870 HD4870)
• RADEON HD 4850 with standard parameters (hereinafter HD4850)
• RADEON HD 3870 X2 with standard parameters (hereinafter HD3870X2)
• RADEON HD 3870 with standard parameters (hereinafter HD3870)
• Nvidia Geforce GTX 260 with standard parameters (hereinafter GTX260)
• Nvidia Geforce 9800 GTX with standard parameters (hereinafter GF9800GTX)

To compare the results of the new RADEON HD 4870 video card, these models of video cards were chosen for the following reasons: it will be interesting to compare it with the RADEON HD 3870 X2, as with a two-chip solution from AMD based on the previous architecture's GPU, in order to evaluate the impact of architecture improvements and the difference in performance. The comparative performance of the RADEON HD 4850 is interesting in order to find out the contribution of the increased GPU frequencies and the use of a new type of GDDR5 memory. Geforce 9800 GTX, although not a direct competitor, is as interesting as the previous generation of Nvidia chips, and the price of the HD 4870 is not so far from its accelerated version of GTX +. And the Geforce GTX 260 is already a direct competitor to the RADEON HD 4870, this comparison will be the main battle.

Direct3D 9: Pixel Filling Tests

The test determines the peak texel rate in FFP mode for a different number of textures applied to one pixel:

Nothing new or interesting, everything corresponds to the difference in frequencies. As usual, video cards do not reach theoretical values. The results of synthetics do not reach theory, the HD 3870, based on the RV670, is the closest to them. But for all new video cards from Nvidia and AMD, the theoretical maximum is not reached in this test. In our test, the RV770 selects about 26-27 texels per clock from 32-bit textures with bilinear filtering, not reaching 40 theoretical ones. The efficiency of Nvidia cards is even lower - 35-37 texels per clock with a theoretical 64.

As for the comparison of the HD 4870 with the direct competitor to the GTX 260, they are very close in this test, but both of them fall short of the GeForce 9800 GTX. The new AMD card is significantly ahead of the old one, and outperforms the younger model of the HD 4800 line in accordance with the frequencies. Interestingly, in the test with one texture, the HD 4870 lags slightly behind the HD 3870, this is due to the theoretically higher performance of the ROP units in the latter with a 32-bit framebuffer without antialiasing. In the case of a large number of textures per pixel, the ability of ROP units does not prevent the card based on RV770 from showing better results. Let's look at the results in the fill rate test:

The second synthetic test measures the fill rate, and in it we see the same situation, but taking into account the number of pixels written to the frame buffer. In cases with 0 and 1 textures applied, the RADEON HD 4870 still has the same slightly lower result than the HD 3870, which is due to the operating frequency of the ROP units. But, as in the previous diagram, in situations with a lot of textures per pixel, the new video card comes out ahead.

Direct3D 9: Geometry Processing Speed ​​Tests

Let's take a look at a couple of extreme geometry tests, and first we will have the simplest vertex shader showing the maximum throughput across triangles:

All modern chips are based on unified architectures, their universal execution units in this test are busy only with geometrical work, and solutions show good results, which clearly rests not on the peak performance of unified units, but on the performance of other units, for example, triangle setup.

The results show this - RV670 and RV770 are very close at similar frequencies. The results of AMD solutions are traditionally higher than those of Nvidia cards. The RADEON HD 4870 outperforms both Nvidia cards and its counterparts in this test. Since we have removed from consideration the intermediate tests for the processing speed of geometry with one light source, we proceed to consider the most complex geometric problem with three light sources, including static and dynamic transitions:

In this variant, the difference between AMD and Nvidia's solutions is better visible, the gap widened slightly, video cards from the second company “sagged”. HD 4870 and HD 3870 are approximately equal at similar frequencies, they are again limited by something like triangle setup, since the numbers have hardly changed since the last test.

Again, all AMD video cards outperform the Geforce 9800 GTX and GTX 260. In real applications, the universal shader processors are busy mainly with pixel calculations, and we are now going to examine their performance.

Direct3D 9: Pixel Shaders Tests

The first group of pixel shaders that we are considering is very simple for modern video chips; it includes various versions of pixel programs of relatively low complexity: 1.1, 1.4, and 2.0.

Although the tests are too simple for modern architectures and do not show their true strength, they are interesting to watch when changing architectures. In simple tests, the performance is limited by the speed of texture fetching, and in the RV770 chip the texturing performance is just improved. This made it possible to achieve victory on all fronts, the HD 4870 outperforms both Nvidia cards in all the considered tasks and is sometimes up to two times faster than the HD 3870.

In more complex tests, the RADEON HD 4870 also shows excellent results, significantly outperforming its predecessor and competitors. But the Geforce GTX 260 is not impressive due to the lower texturing speed, slightly outperforming the 9800 GTX in only two of the most difficult tests. Let's look at the test results of more complex pixel programs of intermediate versions:

Great result for the RADEON HD 4870! In the highly dependent texture rendering test Water, which uses dependent texture fetching from large nesting levels, and maps are ranked by texturing speed, the new model significantly outperforms both Nvidia cards, and the difference with the HD 3870 is simply striking.

The second test is more intensive in computing units, and it is better suited for AMD architectures with a large number of stream processors. In it, the new AMD solution again shows the best result, faster and Geforce GTX 260 and 9800 GTX by 1.5-2 times! And again, compared to the previous generation, the new board has more than doubled speed. The difference with the HD 4850 corresponds to the difference in GPU frequencies.

Direct3D 9: Tests of Pixel Shaders New Pixel Shaders

These DirectX 9 pixel shader tests are even more complex and fall into two categories. Let's start with the simpler shaders version 2.0:

• Parallax Mapping- the texture mapping method familiar from most modern games, described in detail in the article
• Frozen Glass- complex procedural texture of frozen glass with controllable parameters

There are two variants of these shaders: with a focus on mathematical calculations, and with a preference for fetching values ​​from textures. Consider mathematically intensive options that are more promising from the point of view of future applications:

These are mathematical tests that depend on the frequency of shader units and texturing speed, the balance of the chip is important here. The performance of video cards in the Frozen Glass test is limited not only by mathematics, but also by the speed of texture fetching, so the old RADEONs show the weakest result. But the new ones ... See for yourself, they are noticeably faster than the previous one. And the HD 4870 we are reviewing is even ahead of the GeForce 9800 GTX and GTX 260.

In the second test "Parallax Mapping", the new products from AMD are even stronger. While the HD 4850 scores slightly above the GTX 260, the HD 4870 is well ahead of both Nvidia models. Improvements in the TMU significantly strengthened the results of the HD 4800 line, in these tests they became the new leaders. Let's consider the same tests in a modification with a preference for samples from textures to mathematical calculations, where the results can turn out to be even more interesting:

The results of the RADEON HD 4850 and Geforce 9800 GTX are quite close, but the HD 4870 is expected to outperform both due to the higher frequency of the chip. The relative position of the cards has changed slightly, the emphasis on the speed of texture units is noticeable. And both cards based on RV770 outperform the previous top one by two or more times. But the GTX 260 showed very weak results in this case, even lagging behind its predecessor.

Consider the results of two more pixel shader tests - version 3.0, the most difficult of our pixel shader tests for Direct3D 9. The tests differ in that they heavily load ALUs and texture units, both shader programs are complex, long, and include a large number of branches:

• Steep Parallax Mapping- a much more "heavy" version of the parallax mapping technique, also described in the article
• Fur- procedural shader rendering fur

AMD's new architecture shows its best in these tests, in contrast to previous solutions, which were outperformed by Nvidia cards. HD 4870 outperforms all competitors by a large margin, the difference with HD 3870 is simply enormous. And Geforce 9800 GTX with Geforce GTX 260 are far behind.

Once again, we see excellent results from the redesigned AMD architecture in our DirectX 9 benchmarks. But what will happen in DX10, because in past studies, things there were clearly worse. Now let's find out by comparing it with the previous generation dual-chip card, since everything is clear with the single-chip RV670 for a long time ...

Direct3D 10: PS 4.0 Pixel Shader Tests (Texturing, Loops)

The new version of RightMark3D 2.0 includes two familiar PS 3.0 tests for Direct3D 9, which were rewritten for DirectX 10, as well as two more completely new tests. The first pair adds the ability to enable self-shadowing and shader supersampling, which additionally increases the load on video chips.

These tests measure the performance of executing pixel shaders with loops, with a large number of texture samples (in the heaviest mode, up to several hundred samples per pixel!) And a relatively low ALU load. In other words, they measure the texture sampling rate and branching efficiency in a pixel shader.

The first pixel shader test will be Fur. At the lowest settings, it uses 15 to 30 texture samples from the heightmap and two samples from the main texture. Effect detail - “High” mode increases the number of samples up to 40-80, enabling “shader” supersampling - up to 60-120 samples, and “High” mode together with SSAA differs in maximum “severity” - from 160 to 320 samples from the height map.

Let's first check the modes without supersampling enabled, they are relatively simple, and the ratio of the results in the “Low” and “High” modes should be approximately the same.

Performance in this test depends not only on the number and speed of TMUs, but also on fill rate and memory bandwidth. As we expected, in Direct3D 10 tests of procedural rendering of fur with a large number of texture fetches, nothing much has changed - the same huge advantage of Nvidia solutions over AMD. Let's see what happens next, AMD cards always fail this test.

Although the HD 4870 lost to both Nvidia cards, it showed an advantage over the younger model of the line, corresponding to the frequency difference. And the dual-GPU RADEON HD 3870 X2 outperformed the new HD 4870 only in the heavy mode. A very good result if you don't look at the Nvidia numbers. Let's look at the result of the same test, but with the enabled "shader" supersampling, which increases the work by four times, perhaps in such a situation something will change, and the memory bandwidth with fill rate will affect less:

Enabling supersampling theoretically quadruples the load, this time the overwhelming advantage of Nvidia cards has not gone anywhere either, although the new AMD video cards are already clearly closer to the Geforce 9800 GTX. As for the rest, with the increase in the complexity of the shader and the load on the video chip, the difference between the HD 4870 and the dual-GPU HD 3870 X2 is almost the same, they are close to each other.

The second test that measures the performance of complex pixel shaders with loops with a large number of texture fetches is called Steep Parallax Mapping. At low settings, it uses 10 to 50 texture samples from the heightmap and three samples from the main textures. When you turn on the heavy mode with self-shadowing, the number of samples doubles, and supersampling increases this number four times. The most complex test mode with supersampling and self-shadowing selects from 80 to 400 texture values, that is, eight times more than the simple mode. We first check the simple options without supersampling:

This test is more interesting from a practical point of view, because varieties of parallax mapping have long been used in games, and heavy options like our steep parallax mapping are used in some projects, for example, in Crysis and Lost Planet. In addition, in our test, in addition to supersampling, you can enable self-shadowing, which increases the load on the video chip by about two times, this mode is called "High".

The relative position of the cards from the previous test is repeated. Although AMD solutions were strong in Direct3D 9 parallax mapping tests, in the updated D3D10 version without supersampling they cannot cope with our task at the level of Geforce video cards, and enabling self-shadowing causes too much performance drop on AMD products. The RADEON HD 4870 we are reviewing lags behind both Geforce video cards and is very close to the dual-GPU HD 3870 X2. Let's see what will change the inclusion of supersampling, in the last test it caused a greater drop in speed on Nvidia cards.

When supersampling and self-shadowing are turned on, the task becomes more difficult, the combination of two options at once increases the load on the cards by almost eight times, causing a large drop in performance. The difference between the speed of different video cards is already different, the inclusion of supersampling affects the same as in the previous case - AMD cards improve their performance relative to Nvidia solutions. And although the new HD 4800s continue to lag behind Geforce, the HD 4870 is close to the HD 3870 X2 and almost caught up with at least the Geforce 9800 GTX. Of course, it is far from a direct competitor to the GTX 260.

Direct3D 10: PS 4.0 Pixel Shader Tests (Compute)

The next couple of pixel shader tests contain the minimum number of texture fetches to reduce the impact of TMU performance. They use a large number of arithmetic operations, and they measure exactly the mathematical performance of video chips, the speed of execution of arithmetic instructions in a pixel shader.

The first math test is Mineral. This is a complex procedural texturing test that uses only two texture data samples and 65 instructions like sin and cos.

When analyzing the results of our synthetic tests, we always note that in computationally complex tasks, modern AMD architectures perform better than competitors from Nvidia. And now Mineral HD 4870 has simply tore the competitors apart. The top video card based on one RV770 chip outperforms the previous generation card based on two RV670s, which is close to the difference in the number and frequency of stream processors. Also, the new video card is almost twice as fast as the direct competitor Geforce GTX 260, not to mention the Geforce 9800 GTX.

The second shader computation test is called Fire, and it is even harder for ALUs. It has only one texture fetch, and the number of instructions like sin and cos has been doubled, to 130. Let's see what has changed with increasing load:

In this test, the rendering speed is limited solely by the performance of shader units, and the test suits AMD architectures very well, which is clearly noticeable after fixing a bug in AMD drivers. What can I say ... Complete defeat of Nvidia solutions. Just think, RADEON HD 4870 is more than twice as fast as Geforce GTX 260 and faster than dual-GPU HD 3870 X2. An amazing result, the RV770 is clearly the strongest GPU in general. By the way, the speed ratio between HD 4870 and HD 4850 exactly matches the difference in frequencies.

Direct3D 10: Geometry Shader Tests

The RightMark3D 2.0 package contains two tests of the speed of geometry shaders, the first version is called "Galaxy", the technique is similar to the "point sprites" from previous versions of Direct3D. It animates a particle system on a GPU, a geometry shader creates four vertices from each point, forming a particle. Similar algorithms should be widely used in future DirectX 10 games.

Changing the balancing in the geometry shader tests does not affect the final rendering result, the final image is always exactly the same, only the scene processing methods change. The "GS load" parameter determines in which shader the calculations are performed - in vertex or geometric. The number of calculations is always the same.

Let's consider the first variant of the Galaxy test, with calculations in a vertex shader, for three levels of geometric complexity:

The ratio of speeds with different geometric complexity of scenes is approximately the same, the performance corresponds to the number of points, with each step the FPS drop is about two times. The task for modern video cards is not very difficult and the limitation of the speed by the power of stream processors in the test is not obvious, the task is also limited by the memory bandwidth and fill rate.

Well, it turned out very interesting, the dual-GPU HD 3870 X2, the new HD 4870 and the competitor GTX 260 have extremely dense results. And the HD 4850 is very tightly paired with the GeForce 9800 GTX. Interesting ... Perhaps, when transferring part of the calculations to the geometric shader, the situation will be even more interesting, let's see:

But no, the difference between the considered test variants is small, there were no significant changes. Is that the dual-chip HD 3870 X2 came out on top in terms of the achieved frame rate. It's easier for him, the AFR multi-chip rendering algorithm forgives a lot. Nvidia video cards show identical results when changing the GS load parameter, which is responsible for transferring part of the calculations to the geometry shader, and the results of some AMD video cards have slightly increased. Let's see what will change in the next test, which assumes a heavy load on geometry shaders ...

Hyperlight is the second geometry shader test that demonstrates the use of several techniques at once: instancing, stream output, buffer load. It uses the dynamic creation of geometry using rendering in two buffers, as well as a new Direct3D 10 feature - stream output. The first shader generates the direction of the rays, the speed and direction of their growth, this data is placed in a buffer that is used by the second shader for rendering. For each point of the ray, 14 vertices are built in a circle, up to a million output points in total.

A new type of shader program is used to generate "rays", and with the "GS load" parameter set to "Heavy" - also for their rendering. That is, in the "Balanced" mode, the geometry shaders are used only for creating and "growing" rays, the output is carried out using "instancing", and in the "Heavy" mode, the geometry shader is also involved in the output. Let's look at light mode first:

Relative results in different modes correspond to the load: in all cases the performance scales well and is close to the theoretical parameters, according to which each next level of "Polygon count" should be twice as slow. This time the RADEON 4850 and HD 4870 are faster than the dual-GPU solution based on the previous GPU architecture, but all AMD cards lag behind all Nvidia solutions, although the HD 4870 is close to them.

It seems that the results of the new maps were influenced by the improved texturing capabilities. However, the numbers should change in the next diagram, in a test with more active use of geometry shaders. It will also be interesting to compare the results obtained in "Balanced" and "Heavy" modes with each other.

This time only the GeForce 9800 GTX failed, all other architectures withstood the blow. In both the RV770 and GT200, some optimizations have been made to improve geometry shader performance. And the RADEON HD 4870 has now caught up with the Geforce GTX 260, except for the simplest mode. The previous generation of AMD chips performs significantly worse in this test, even the dual-GPU video card lags behind.

As for the comparison of results in different modes, everything is as always, AMD video cards improve their performance when switching from instancing to a geometric shader during output, and old Nvidia video cards lose performance. A Geforce card based on the G92 chip can compete only due to the speed in the "Balanced" mode, which is almost equal to the speed in the "Heavy" of the RADEON HD 4850. At the same time, the picture obtained in different modes does not visually differ.

Direct3D 10: Speed ​​of fetching textures from vertex shaders

The Vertex Texture Fetch tests measure the speed of a large number of texture fetches from the vertex shader. The tests are similar in essence and the ratio between the results of the cards in the tests "Earth" and "Waves" should be approximately the same. In both tests, it is used based on texture samples, the only significant difference is that the Waves test uses conditional transitions, while the Earth does not.

Consider the first test "Earth", first in the "Effect detail Low" mode:

Judging by previous studies, the results of this test are influenced not only by the texturing speed, but also by the ROP performance and memory bandwidth, and the simpler the mode, the more impact on the speed they have. In all modes, except for the simple one, the leader is the top model of the HD 4800 series, which we are considering today. In idle mode, memory bandwidth affects, and multi-chip rendering shows itself quite well. GTX 260 only shows results at the HD 4850 level. Let's look at the results of the same test with an increased number of texture fetches:

The situation has not changed too much, but texturing has a stronger effect on speed, as can be seen from a pair of Geforce. HD 4870 has lost ground and is not a leader, although it lags just a little behind the GeForce 9800 GTX in difficult modes. In idle time, the GTX 260 with a large memory bandwidth is in the lead. Interestingly, as the geometry becomes more complex, the difference between HD 4870 and HD 3870 X2 changes.

Let's consider the results of the second test of texture fetching from vertex shaders. The Waves test has a smaller number of samples, but it uses conditional jumps. The number of bilinear texture samples in this case is up to 14 ("Effect detail Low") or up to 24 ("Effect detail High") for each vertex. The complexity of the geometry changes in the same way as in the previous test.

The second test of this section called "Waves" is more favorable to AMD products; the new model of the RADEON HD 4800 family looks very good, at the level of its two-GPU predecessor. And it also outperforms Nvidia graphics cards, except for the simplest, where the GTX 260 is slightly ahead. It seems that under such conditions the TMU efficiency of the RV770 is higher than that of the Nvidia GPU. Let's consider the second variant of the same test:

And again we see very little new, although with an increase in the complexity of the test, the results of AMD video cards relative to the speed of Nvidia cards improved, the latter lost a little more due to changes in testing conditions. In the lightest mode, HD 3870 X2 and HD 4870 are ahead, in the rest the dual-GPU HD 3870 X2 has no equal. Well, among single-chip cards the hero of the review is the best, he is ahead of his younger brother HD 4850 according to the difference in frequencies. Nvidia cards are left behind this time.

3DMark Vantage: Feature tests

We decided to include synthetic tests from 3DMark Vantage into the RADEON HD 4870 review. The package is new, its feature tests are quite interesting and differ from ours. Probably, when analyzing the results of the maps in this package, we will draw some new and useful conclusions for ourselves.

Feature Test 1: Texture Fill

The first test is the texture sampling speed test. Used to fill a rectangle with values ​​read from a small texture using multiple texture coordinates that change every frame.

The balance of results is broadly similar to what our tests show, using conditions in which Nvidia cards do not benefit from a large number of TMUs. The old single-GPU card from AMD lags far behind everyone, but the dual-GPU HD 3870 X2 and the new HD 4870 model significantly outperform both solutions from Nvidia. Geforce GTX 260 lags behind Geforce 9800 GTX, as it should be in theory. But why does the RV770-based card outperform both the G92 and the GT200? Apparently, the point is in the very efficiency of texture units, which is higher in AMD cards.

Feature Test 2: Color Fill

Fill rate test. A very simple pixel shader is used with no performance limitation. The interpolated color value is written to an offscreen buffer (render target) using alpha blending. A 16-bit FP16 off-screen buffer is used, which is most often used in games that use HDR rendering, so this test is very timely.

The readings of this test correspond to what we get in our synthetic tests, taking into account the fact that we use an integer buffer with 8-bits per component, and in the Vantage test - 16-bit floating point. Therefore, all the numbers are half ours.

These numbers rather show not only the ROP performance, but also the amount of memory bandwidth (in the case of multichips, multiplied by the number of chips for AFR). The figures correspond to the theoretical ones and depend, first of all, on the width of the memory bus and its frequency. In this test, the new HD 4870 model, taking advantage of the improved ROP capabilities and large GDDR5 memory bandwidth, performs better than the dual-GPU HD 3870 X2 and GTX 260 with 448-bit memory bus.

Feature Test 3: Parallax Occlusion Mapping

One of the most interesting feature tests, as a similar technique is already used in games. It draws one quadrangle (more precisely, two triangles), using a special technique Parallax Occlusion Mapping, which simulates complex geometry. Quite resource-intensive ray tracing operations and high resolution depth map are used. Also, this surface is shaded using the heavy Strauss algorithm. This is a test of a very complex and GPU-heavy pixel shader containing numerous texture selections for ray tracing, dynamic branches and complex lighting calculations using Strauss.

The test is interesting in that it does not depend only on shader power, branching efficiency and texture sampling rate, but on everything at once. That is, to achieve high speed, the balance of the chip and the card is important. And the most important thing is the efficiency of branching in shaders, the so-called execution granularity.

The old cards from both manufacturers are far behind, even the dual-GPU HD 3870 X2 could not catch up with the HD 4870, although the dual-GPU rendering of this test is quite efficient. And here we see an interesting arrangement of the RADEON HD 4870 and Geforce GTX 260. Despite the fact that in the tests of texture sampling and mathematical calculations the AMD solution usually wins, in the POM test the Geforce is stronger than the RADEON. And the best efficiency of processing branches in shaders in GT200 is to blame for this.

Feature Test 4: GPU Cloth

The test is interesting in that it calculates physical interactions (imitation of tissue) using a video chip. Vertex simulation is used, using combined work of vertex and geometry shaders, with several passes. Use stream out to transfer vertices from one simulation pass to another. Thus, the performance of vertex and geometry shaders execution and stream out speed are tested.

In this test, we traditionally get strange results with dual-GPU cards, the HD 3870 X2 does not get any acceleration from its second GPU. As for the rest, we again see the lag in AMD solutions, even the relatively fast HD 4870 falls short of the Geforce 9800 GTX, not to mention the GTX260. It seems that the speed does not depend on the shader performance, but depends on the stream out speed ...

Feature Test 5: GPU Particles

Physical simulation test of effects based on particle systems calculated using a video chip. Vertex simulation is also used, each vertex represents a single particle. Stream out is used for the same purpose as in the previous test. Several hundred thousand particles are calculated, all are animated separately, and their collisions with the height map are also calculated. Similar to one of the tests in our RightMark3D 2.0, particles are rendered using a geometry shader, which creates four vertices from each point, forming a particle. But the test most of all loads shader units with vertex calculations, stream out is also tested.

Here we see almost the same thing as in the previous case, only Geforce 9800 GTX lags behind, and AMD cards moved up higher. But still, Geforce GTX 260 remains the leader, today's hero HD 4870 is close to it. The dual-GPU card from AMD again did not go far from the old one-GPU card, and both are at the bottom of the list. And again, suppose the speed is affected by stream out performance, memory bandwidth and texture performance at the same time.

Feature Test 6: Perlin Noise

This feature test can be considered a mathematically intensive test of a video chip, it calculates several octaves of the Perlin noise algorithm in a pixel shader. Each color channel uses its own noise function for more load on the video chip. Perlin noise is a standard algorithm often used in procedural texturing and is very complex mathematically.

The latest feature test in Vantage shows the pure mathematical performance of video chips. The performance shown in it is quite consistent with what we see in our math tests from RightMark 2.0. AMD video cards naturally outperform their competitors from Nvidia, even the dual-GPU HD 3870 X2 outperforms the GTX 260. And the RADEON HD 4870 is the leader and outperforms its main competitor by more than 25%.

Conclusions on synthetic tests

Based on the results of the synthetic tests carried out, we confirm the conclusions made in the previous article. New solutions from AMD turned out to be very successful, many changes were made in the RV770 chip, in almost all synthetic tests it is several times faster than the previous generation video card. Thanks to the improved architecture of the RV770, which corrected the main drawbacks, the RADEON HD 4870 outperforms its main competitor Geforce GTX 260 in many tests. The RV770 has become more efficient and balanced, which is important for modern and future 3D applications with a large number of complex shaders.

The RV770 chip has a large number of execution units, supports the new GDDR5 memory, which made it possible to release the RADEON HD 4870 with a high memory bandwidth based on only a 256-bit memory bus. Small questions arise only about the lower efficiency of branching in shader programs, which affects the performance of the most complex parallax mapping algorithms. Well, in terms of stream out speed, new solutions from AMD are inferior to competitors from Nvidia. The rest of the new HD 4800 line is just fine! Especially the computational performance, in which they are far ahead.

The next part of the article contains tests of the new solution from AMD and other video cards in modern gaming applications. The game results should confirm our findings. It can be assumed that the average speed of the HD 4870 in games will be approximately on par with the Geforce GTX 260.

The power supply for the test bench was provided by the company TAGAN
Dell 3007WFP Testbed Monitor Courtesy of

Each brand in the electronics market has one unique product that brought fame to its manufacturer and went down in history due to its features. ATI's flagship is the Radeon HD 4870 gaming video adapter. The review, characteristics, photos, reviews of the owners in this article will allow the reader to get to know this interesting product better.

Current state of affairs

It should be noted right away that the once popular and very high-performance video card has been occupying the last positions in the rating of gaming devices for several years, however, as practice shows, the graphics accelerator is not going to give up without a fight. If the reader looks at the system requirements for the latest games presented on the market a few months ago, he will certainly see that for a comfortable game, the developers recommend using the Radeon HD 4870, the characteristics of which meet the minimum requirements.

It is clear that this video adapter is unlikely to be suitable for playing at high resolutions in a resource-intensive application, but the graphics accelerator still has the potential, especially when it comes to a dual-core market representative. But first things first.

Entry level gaming class

The 4870 video card, unlike other products of ATI's 4000 series of video adapters, received a new RV770 core, and, accordingly, a modern set of instructions with support for all technologies existing at that time. The manufacturer has managed to accommodate almost 1 billion transistors on a 256 mm 2 die area. The graphics accelerator has 800 universal processors, 16 rasterization units and 40 texture units. This video card was one of the first to receive an advanced 256-bit GDDR5 bus operating at 900 MHz.

The factory-made graphics core has received an underestimated frequency equal to 750 MHz. However, the reader should not pay attention to this indicator, since the final products released under well-known brands have solved the issue with the speed of the GPU. It is better to pay attention to the power consumption of the gaming adapter - 160 watts at maximum load. There is something to think about for all owners of low-power power supplies.

All known technologies

Since performance characteristics are in the first place for the Radeon HD 4870 video adapter, the manufacturer simply had to provide his product with the DirectX 10.1 libraries that were up-to-date at that time. The OpenGL standard is limited to 3.1, which greatly confuses potential buyers, because at the time of the presentation of the device, OpenGL 4.0 was already announced. The most interesting thing is that the device has received support for new shaders version 4.1. It is this technology that still allows the video card to remain in demand in the gaming market. The video adapter works with the outdated PCEex16 version 2.0 interface. The maximum degree is 16x.

Also important for the graphics accelerator is support for the popular DisplayPort and HDMI video outputs. The audio controller integrated on the video adapter also supports the transmission of 7-channel audio. Surprisingly, this technology has been implemented on discrete platforms for both personal computers and mobile devices. It should be noted that the analog video output is limited to a resolution of 2048x1536 dpi, while this parameter is no better for the digital interface - 2560x1600 dpi.

Double punch

After conducting the first tests of the gaming video adapter in laboratory conditions, the technologists came to the conclusion that the Radeon HD 4870 video card, the characteristics of which can be modified, could become the flagship of ATI. The changes affected primarily the power supply system and phase distribution. The manufacturer also installed two RV770 graphics cores on the video adapter. As a result, the market saw a new champion in benchmark and gaming performance - the HD 4870 X2 graphics card.

All characteristics of the nominal device have actually doubled. Even in laboratory tests, the video adapter showed identical results to the base device operating in CrossFire X2 mode. The only difference that should be attributed to the merits is the reduction in power consumption of the device. The manufacturer managed to reduce the consumption of two cores to 280 watts. It looks pretty weird. After all, logically, the manufacturer simply had to rebrand single-core chips with high power consumption.

Strike while the iron is hot

Notebook manufacturers are also interested in the gaming graphics core, so ATI has presented to the market a mobile solution based on a new graphics core. Since for the Radeon HD 4870 chip, temperature and power consumption are critical indicators, the manufacturer was forced to reduce the frequency of the GPU to 550 MHz. But even these indicators are enough to provide a comfortable game for many users.

All parameters, including the technologies used, migrated to the mobile platform. Naturally, the 256-bit GDDR5 bus with a volume of 1 GB did not go unnoticed by many potential buyers. As you might expect, many laptop manufacturers have not paid enough attention to the cooling system, which is why they have attracted a lot of negativity from the owners of mobile devices. Due to massive user dissatisfaction, manufacturers soon abandoned the Radeon HD 4870.

Direct competitor

AMD started a vigorous activity to create a gaming solution after the competitor Geforce 9800 GT appeared on the market. The Radeon HD 4870 was supposed to move the market leader in the gaming segment. But, as often happens, the company's technologists overdid it and surpassed the performance of the Nvidia graphics accelerator. Many experts in their reviews assure others that this particular chip is the link between the two generations of gaming video cards.

And if we return to the competition, then we can safely say that the Radeon HD 4870 video card has accelerated the emergence of new productive devices in the gaming market. After all, the Nvidia GTX 260 graphics accelerator became the first representative of a new generation of video adapters, which was supposed to surpass all competitors in performance.

Appearance and quality

A plastic protective case and a turbofan on the Radeon HD 4870 graphics accelerator (a photo of the device can be seen in this article) indicate that the buyer has a representative of the gaming segment. the video card is made using red PCB, and the heatsink cover and high-frequency cooler have a bright scarlet color to match the AMD product.

With a detailed study of the placement of microcircuits on a graphics accelerator, it will become clear to any user that the manufacturer will not sell its products at retail, but will provide this opportunity to the leaders of the computer market. Naturally, the brand that can provide a device with powerful heat dissipation with a decent cooling system will win in the performance race.

Chasing the stars

It is better to start with the Radeon HD 4870 graphics accelerator with the most productive products from HIS and Sapphire, which used the 4870 X2 chip with two cores to create gaming solutions. Many owners in their reviews note that at the time of the announcement, these video cards simply had no competitors in the market for gaming components for a computer. In addition to high performance, video adapters have many other advantages:

• a unique package including a set of necessary adapters and interface cables;
• completely redesigned PCB power system with the addition of several phases;
• replacement of electrolytic capacitors with solid-state capacitors;
• a self-produced cooling system was installed with efficient heat removal from both the graphics core and memory chips.

Returning to the HIS product package, it is worth noting that the manufacturer has equipped its video card with a proprietary device that combines a screwdriver, a level and a flashlight. Also in the box, the buyer will find a coupon for free download of any licensed game from the Half-Life series.

Affordable product

The best-selling video adapter based on the Radeon HD 4870 chip, with which all enthusiasts of the world preferred to compare, is a product from ASUS. The manufacturer managed to reach the golden mean in the price-performance criterion, forcing many representatives of the entry-level gaming segment to lower the cost of their products in order to somehow stay on the market.

The main feature of the video card is the developer's increase in the number of phases for powering the printed circuit board circuits. Also, technologists have developed an effective open-type cooling system. The core and heatsink of the cooler are made of pure copper and are mounted in the device only on the graphics core. The memory chips (the volume of which has been halved to only 512 MB) are blown by the fan superficially and do not have their own heatsinks for heat dissipation.

Gaming performance and overclocking potential

Many owners in their reviews note that the performance characteristics of the Radeon HD 4870 1 Gb video card directly depend on the operation of the graphics core. Accordingly, for overclocking, you need to worry about the presence of a powerful cooling system. All products with HD 4870 chip have some problems with memory modules. No brand has been able to overcome the 1000 MHz barrier (4000 MHz effective frequency). One conclusion suggests itself - the printed circuit board has problems when working with fast memory. There is only one advantage - the memory chips do not heat up.

The buyer can safely count on the performance of the gaming adapter with games such as Crysis 4, World of Tanks and their counterparts, which share a common rather than focusing on the potential of the graphics accelerator. As practice shows, it is better to install the Radeon HD 4870 video card with a low-frequency dual-core or 4-core processor.

Driver problems and solutions

After the updated Windows 10 system entered the world market, many game lovers had problems installing drivers. The official website of the manufacturer lacks software support for all devices based on the 4000 series of video cards. Once again, AMD, after updating the operating system, refuses to support its products (competitor Nvidia has much higher loyalty to the owners of old devices).

To correctly connect to the Radeon HD 4870 video card at the software level, you need to use Windows Update. There is no need to expect that the full package of utilities will be downloaded, you just need to download it from the manufacturer's official website. Microsoft will provide only one driver that will make the graphics accelerator compatible with the motherboard when connected. This approach of AMD to customers causes a lot of negativity from the owners of all Radeon-based devices.

With the recent release of the Radeon HD 4000 series, video card prices have dropped; now a very solid gaming solution can be purchased for $ 250.

A month ago we reviewed the fresh Radeon HD 4850, an inexpensive graphics card from ATI's new line, and fell in love with it. After that, Nvidia revised its prices - now for $ 250 you can buy both the Radeon HD 4850 and the GeForce 9800 GTX (they used to ask for more than $ 330 for it).

However, if you can afford to spend a little more, AMD offers the Radeon HD 4870, a single-core graphics card with even more performance for around $ 380. In the same price category are the Radeon HD 3870 X2 and GeForce GTX 260; prices for GeForce 9800 GX2 and GeForce GTX 280 are much higher.

Today we will take a look at the Visiontek Radeon HD 4870 video card and compare it with all the above-mentioned top-end cards. How is it in terms of price-performance ratio?

The characteristics of the Radeon HD 4870 are very impressive. The GPU consists of 800 stream processors (160x5), 40 texture units and 16 rasterization units. Compare these parameters with those of the previous generation Radeon HD 3870 - which boasts only 320 stream processors (64x5), 16 texture units and 16 ROPs - and you can imagine how the numbers have changed with the release of the new ATI line. The memory bandwidth of the Radeon HD 4870 is 115.2 GB / s, while the Radeon HD 3870 is capped at 57 GB / s.

This leap in bandwidth is due to the use of the latest GDDR5 memory operating at 900 MHz (1.8 GHz x 2 = 3.6 Gbps). The core operates at 750 MHz (1.2 TFlops), 125 MHz faster than the HD 4850 core. Apart from the differences in memory characteristics and frequencies, the HD 4870 and HD 4850 are almost identical.

The core of the Radeon HD 4870 is built using 55nm technology and includes a 256-bit memory bus. Like the HD 3000 series, the HD 4000 series uses a PCI-Express 2.0 slot, but is compatible with older PCI-Express 1.0 slots as well. Unlike the latest GeForce cards, the Radeon HD 4000 series fully supports DirectX 10.1, which may come in handy later on. The Radeon HD 4870 will ship with either 512 or 1024 MB of memory on board; our today's sample is equipped with 512 MB of GDDR5 memory.

Video card

An early market launch can play a joke with the Visiontek card - it is very similar (for us, testers, of course) to the reference card from AMD itself. The only visible difference is the Visiontek label on the cooler. In any case, the video card looks impressive, especially the cooling system that occupies two slots.

Like the core of the Radeon HD 3870, the core of the new Radeon HD 4870 is manufactured using 55nm technology. Despite this, ATI didn’t try hard to overclock - the core runs at 750 MHz. The GDDR5 memory runs at a more impressive 3.6 GHz (1.8 GHz x2) (on this particular model). As we mentioned above, the Radeon HD 4870 carries 800 stream processors along with 40 texture units and 16 raster units. Summing up, we can safely say that these are outstanding characteristics for a video card costing $ 350-380.

Visiontek ships its Radeon HD 4870 with Qimonda IC memory on board (IDGV51-05A1F1C-40X). These GDDR5 modules are stable at 4.0GHz (2.0GHz x2), so further overclocking should be relatively easy.

The Visiontek Radeon HD 4870 is designed with a dual-slot cooling system. The big advantage of this solution is a very quiet cooler. But there are also disadvantages, which directly follow from the advantages: in idle mode, the card heats up to 74 degrees, which is not encouraging, and under load the temperature has risen to almost 82 degrees.

About dimensions. The Radeon HD 4870 is only half a centimeter longer than its fellow Radeon HD 4850; total - 24 cm (9.5 inches). Like other cards in its segment, the Radeon HD 4870 needs two 6-pin power connectors.

The Visiontek card also supports HDMI and 7.1-channel audio via the included adapter; on the card itself you can see two doubly connected DVI ports and an S-Video output. Also included is a DVI-VGA adapter, HDTV component cable and Crossfire bridge.

Tests: Configuring test bench and 3Dmark

Test bench configuration
- Intel Core 2 Quad Q6600 @ 3.00GHz (LGA775)
- x2 Kingston HyperX 2GB PC2-8500
- ASUS P5Q Deluxe (Intel P45)
- OCZ GameXStream (700W)
- Seagate 500GB 7200 rpm (Serial ATA300)
- ASUS GeForce GTX 280 (1GB) - 177.41
- Gigabyte GeForce GTX 260 (896MB) - 177.41
- ASUS GeForce 9800 GX2 (1GB) - 174.74
- ASUS GeForce 9800 GTX + (512MB) - 174.74
- ASUS GeForce 9800 GTX (512MB) - 174.74
- Visiontek Radeon HD 4870 (512MB) - 8.7
- ASUS Radeon HD 4850 (512MB) - 8.7
- VisionTek Radeon HD 3870 X2 (1GB) - 8.7
Software
- Microsoft Windows Vista Ultimate (64-bit)
- Intel System Driver 8.4.0.1016
- Nvidia Forceware 177.34
- Nvidia Forceware 174.74
- Nvidia Forceware 169.44
- ATI Catalyst 8.7

Based on the 3DMark Vantage results, we can assume that the Visiontek Radeon HD 4870 will slightly outperform the GeForce GTX 260 in our next tests, while the GeForce GTX 280 will be much faster.

But we are sure that in fact it will not go exactly that way, so let's start testing in real gaming applications.

Tests: Company of Heroes

Visiontek Radeon HD 4870 in Company of Heroes does not look so impressive - it comes in the penultimate of all tested 8 cards and only outperforms the Radeon HD 4850.On the other hand, at 1920x1200 it is only 1 fps slower than the GeForce 9800 GTX (which, however, at 1680x1050 it breaks off by all 5 fps).

But turning on the 8xAA / 16xAF settings (anti-aliasing and anisotropic filtering) helps the Visiontek Radeon HD 4870 rise into the air - here it turns out to be the third. As you can see from this and previous tests, Company Of Heroes is very sympathetic to Nvidia cards, at least in DirectX 10 mode.

Benchmarks: Crysis

In Crysis, the Visiontek Radeon HD 4870 again does not really show itself, although it catches up with the competitor in the face of the GeForce GTX 260, leaving the GeForce 9800 GTX + slightly behind.

In the DirectX 10 Crysis mode, the results remain about the same: the Visiontek Radeon HD 4870 is on par with the GeForce GTX 260.

Tests: Devil May Cry 4

In our test in Devil May Cry 4, the Visiontek Radeon HD 4870 outperforms the GeForce GTX 260. The ATI card sits between the GeForce GTX 260 and 280, leaving the Radeon HD 4850 far behind.

Enabling 8xAA in Devil May Cry 4 has almost no effect on the results of the Visiontek Radeon HD 4870 (the same can be said for the Radeon HD 4850). HD 4870 beats the more expensive GeForce GTX 280 in 1440x900 and 1680x1050 resolutions, just slightly inferior in 1920x1200.

Tests: Enemy Territory: Quake Wars

If the games Company of Heroes and Crysis prefer cards from Nvidia, then Enemy Territory: Quake Wars is more friendly towards ATI solutions. With anti-aliasing and filtering turned off, the Visiontek Radeon HD 4870 is the fastest here by a small margin. Well, let's see what happens if you enable these options ...

Enabling 8xAA / 16xAF allows the Visiontek Radeon HD 4870 to further expand the lead over the GeForce GTX 280 - the ATI card is undoubtedly the fastest in all three screen resolutions. By the way, in this test the Radeon HD 4850 leaves behind the GeForce GTX 260.

Tests: Supreme Commander

The Visiontek Radeon HD 4870 in Supreme Commander behaves well and keeps up with the GeForce GTX 280 at 1920x1200 and 1680x1050.

With 8xAA, the Radeon HD 4870 lags slightly behind the GeForce GTX 280 at high resolutions. However, the Visiontek video card is still faster than the GeForce GTX 260 in all resolutions.

Tests: Unreal Tournament 3

In Unreal Tournament 3, the Visiontek Radeon HD 4870 shows its best side again. Although both GeForce GTX cards lead in 1440x900 resolution, in higher resolutions the Radeon cards almost catch up with them in this not-so-fresh game.

Enabling anti-aliasing and anisotropic filtering in Unreal Tournament 3 does not help the Visiontek Radeon HD 4870 outperform GeForce GTX cards. While the numbers are pretty high, Unreal Tournament 3 is definitely more sympathetic to the new generation cards from Nvidia.

Tests: World in Conflict

The Radeon HD 4870 in the World in Conflict test gets close to the GeForce GTX 260, but it is beyond the power of Radeon to overtake it without anti-aliasing and anisotropy enabled.

The included 4xAA / 16xAF options allow the Visiontek Radeon HD 4870 to outperform the GeForce GTX 260 at 1920x1200, although things are still the same at 1440x900 and 1680x1050.

Power consumption and heating

The Visiontek Radeon HD 4870, unlike the HD 4850 we tested earlier, is very power hungry and consumes 193 watts in idle mode (this figure is the total power consumption of the system). As a result, even the GeForce GTX 280 is preferable in this mode.

Under load, the Visiontek Radeon HD 4870, on the contrary, is more modest and consumes much less than the GeForce GTX 280, GTX 260, 9800 GX2 and Radeon HD 3870 X2.

The Visiontek Radeon HD 4870 heats up just awfully: even the temperature in idle mode does not drop below 74 degrees. Although under load the large (80mm) cooler was able to keep the card within 82 degrees, we cannot call this state of affairs normal.

Although we did not experience any problems due to such strong heating, it should be noted that the slower Radeon HD 4850 heats up even more. This is easily explained by the fact that its cooler takes up one slot, while the HD 4870 uses two.

We experimented and found out: if the Arctic-Cooling S1 cooler is installed on the Visiontek Radeon HD 4870, the temperature under load will drop to 49 degrees, and in idle mode - to a much more acceptable 33. In other words, buy your new Radeon from the HD series. A third party 4000 cooler is not a bad idea if you want to keep the overall temperature in your computer case as low as possible.

Final thoughts

The Visiontek Radeon HD 4870 is very similar to AMD's reference card, so it is difficult to distinguish it from other Radeon HD 4870 based cards released before June 25th. In any case, this video card does not disappoint and offers simply superb speed for its $ 350-380.

For the same price, you certainly won't find anything better than a new card from the HD 4000 series. Although Nvidia adjusted its prices pretty quickly, we should be grateful to ATI for bringing such a convincing line to the market. The Radeon HD 4850 is still the best choice for budget users, but if you can spend a little more, the Radeon HD 4870 will fully pay off your expectations and money.

As was the case in the Radeon HD 4850 benchmark, there were problems with the reference cooling system used in the Radeon HD 4870. Although it is relatively quiet, the fresh Radeon did get to some extreme temperatures. After installing the cooler Arctic-Cooling S1, the idle temperature dropped from 74 degrees to 33. The temperature under load also dropped significantly - to 52 degrees.

By the way, we were informed that some customers, using third-party software like RivaTuner, were able to set a fixed speed of the reference cooler. Unfortunately, already at half its capacity, this fan becomes too loud.

In addition, with the standard cooler and automatic adjustment of its speed, overclocking was not at all successful. However, using Arctic-Cooling S1 and RivaTuner, we easily raised the core and memory frequencies up to 835 MHz and 4.2 GHz (2.1 GHz х2), respectively.

We really hope that Visiontek intends to present a better solution with a cooler, as it already did with its overclocked Radeon HD 3870 X2. In addition, despite the graphic design of the box with the video card in the style of the Mass Effect game, the manufacturer does not supply the game itself (or any game or game in general) in the kit. Well, at least Visiontek's lifetime warranty is here.

So, if a very hot graphics card is not a problem for you, the Visiontek Radeon HD 4870 will be the best value for your money.

Leadership in the Hi-End class of graphics accelerators for the companies that produce them is a matter of prestige. Despite the relatively low demand for the fastest video cards, the cost of which in the last two or three years is almost always higher than $ 600 or even $ 700, they are a kind of benchmark for most users who are more or less interested in games. It is video cards of this class that demonstrate a kind of reference performance in modern games. And if products based on graphics processors from ATi or from NVIDIA occupy this "throne", then the company itself is considered the leader in a certain period of time, which undoubtedly affects the sales growth of video cards in all price segments and the manufacturer's rating rise.

In my opinion, from the moment the NVIDIA GeForce 8800 GTX / Ultra appeared on the market and up to and including the release of the GeForce GTX 260/280 line, the leadership in the top class of video cards belonged to NVIDIA. For its part, ATi (and later AMD) attempted to regain the first place by releasing a two-chip Radeon HD 3870 X2. However, this was never done in full: problems with the drivers, the AFR rendering mode that had no alternative for ATi, the presence of the GeForce 9800 GX2 on the market, and the new GPU from NVIDIA G200, which arrived very quickly, did not allow HD 3870 X2 to win the fame of the fastest graphics card.

advertising

Nevertheless, having released the certainly successful mid-range video cards Radeon HD 4850 and HD 4870, ATi's graphics division, now owned by AMD, announced on August 12, 2008 a dual-chip Hi-End video card - Radeon HD 4870 X2 with an MSRP of $ 549 and clear claims to absolute leadership. A little later, its younger sister, the Radeon HD 48, should appear on the market. 50 X2 with a recommended value of up to $ 400, which we will also study and test later. Well, today we present to your attention a review and "speed test" of the Radeon HD 4870 X2, released under the Hightech Information System Limited (HIS) label.

1. Review of HIS Radeon HD 4870 X2 2x1 GB

The design style and dimensions of the box in which the HIS Radeon HD 4870 X2 is supplied have not changed in comparison with the earlier reviewed Radeon HD 4850 and HD 4870 from the same manufacturer. Unless the color scheme is now predominantly green:

On the front side of the box, you can find information about the video card model, its interface, volume and type of installed video memory. The reverse side is replete with lines of motherboard specifications, descriptions of system requirements, as well as information about the features of graphics technologies embedded in the GPU. It also lists the awards for print and electronic publications (of which there are already more than 600) received by HIS products. The video card is produced in China.