by Alexey Stepin
04/12/2006 | 11:30 AM
The Radeon X1800 graphics card family is not anymore in the forefront of ATI Technologies’ product range as it has been superseded with the more advanced Radeon X1900 series. Its senior model, Radeon X1800 XT, didn’t come out of the fight with the GeForce 7800 GTX as an absolute champion, although indeed was faster across a number of applications. The later released GeForce 7800 GTX 512 won the battle for Nvidia because the non-overclocked Radeon X1800 XT, even though with a huge core frequency of 625MHz, couldn’t do anything against that card, especially as the GeForce 7800 GTX 512 was equipped with very fast graphics memory clocked at an unprecedented 1700MHz. Still there was an attempt to create a Radeon X1800-based graphics card capable of competing with the GeForce 7800 GTX 512: ASUSTeK Computer came up with a unique version of Radeon X1800 XT that we are going to discuss in this review.
<%BANNER[article]%>Like some other manufacturers’, ASUS’s product range includes graphics cards with frequencies higher than the reference ones, but unlike, for instance, pre-overclocked products from EVGA (for details see our article called EVGA e-GeForce 7800 GTX Graphics Card Review), ASUS’s feature sophisticated cooling systems to ensure highest stability under extreme operating conditions, support a number of exclusive technologies, and are accompanied with truly gorgeous accessories. We already told you about one such card (see our ASUS EN7800GTX TOP/2DHTV review) and praised its superb performance (generally higher than that of the Radeon X1800 XT) as well as its excellent noise characteristics. The Arctic Cooling NV Silencer 5 cooler installed on that card not only made it nearly silent but also efficiently cooled, with hot air being exhausted outside of the PC case. Of course, this came at a price, as usual. That graphics card had a near zero overclocking potential, but that wasn’t such a terrible drawback considering its high performance at default clock rates.
As you know, ASUS turns out graphics cards on ATI’s GPUs, too, so a similar product with the Radeon X1800 chip was sure to appear and has indeed been released. It is called ASUS EAX1800XT TOP/2DHTV and is claimed to be 10% faster than the standard Radeon X1800 XT.
This performance gain is achieved by lifting the card’s frequencies from the standard 626/1500MHz to an impressive 700/1600MHz. Let’s learn more about this card to see if it is really capable of rivaling the formidable GeForce 7800 GTX 512 and how it compares with the new Radeon X1900 XT.
ASUS’s top-end graphics cards are usually packed in a huge box, but the huge box with the EAX1800XT TOP is designed in a different way than before:
ASUS’s graphics cards on ATI’s GPUs used to come in blue-colored rather than black boxes. There is another picture on the box’s front, too. Not a fantasy knight or mage is looking sternly at the buyer, but a space trooper in full armor and with an impressive gun in hand. The box still has a convenient handle and its front panel can be opened for you to read a detailed product specification. Unfortunately, there is now no window that would allow taking a look at the card without taking it out of the package.
Here are the accessories you receive with the ASUS EAX1800XT TOP/2DHTV graphics card:
As you see, none of the necessary cables and adapters is missing here. We’ll talk about the external power supply shortly while the software bundle deserves closer inspection. It includes the following games:
Besides games, the software pack also includes discs with drivers and utilities, with the ASUS DVD XP player, with CyberLink Power Director 3 and Media@Show SE 2.0. An individual CD carries a detailed user manual in digital format. The discs are supplied in traditional paper envelopes, but you can store them in the enclosed imitation-leather pouch which looks quite stylish to us.
There is no web-camera here which used to accompany some earlier graphics cards from ASUS; it is replaced with an ASUS XitePad joystick:
This is a regular 10-button gamepad with USB interface, two analog sticks and feedback. Its rubberized plastic case is quite handy; the buttons and sticks are placed in the same manner as on gamepads from Sony’s PlayStation which are considered among the most ergonomics-friendly devices in their class.
We think it’s really better to receive a gamepad than a web-camera for such a gaming graphics card as EAX1800XT TOP. You are sure to appreciate it if you are into games with third-person view, especially as a fine example of the genre, Peter Jackson’s King Kong , is the card’s software bundle.
The EAX1800XT TOP supports the following exclusive technologies from ASUS:
You will find the technologies-related software on the Drivers CD.
The packaging and accessories of the ASUS EAX1800 XT TOP/2DHTV are really top-notch. We can think of only one card, MSI’s NX6800GT-T2D256E, that comes with as many games included (for details see our article called MSI NX6800GT-T2D256E Graphics Card Review).
ATI Technologies doesn’t permit its partners any experiments with the reference design of its high-end graphics cards, but provides them with ready-made devices manufactured at Sapphire Technologies’ facilities. This is why the PCB of the ASUS EAX1800XT TOP doesn’t differ from the PCB of the standard Radeon X1800 XT. They only differ in the design of the mounting bracket and in the cooling system.
The L-shaped mounting bracket serves as a reinforcing frame here, going along the long side of the PCB to prevent any damage under the weight of the cooler. A power cable, attached to the card’s external power connector, is hidden in the bracket. The other end of the cable ends in a round connector which you can see in the card’s back above the first DVI-I output. This connector is meant for an external 80W power supply.
Yes, this graphics card comes with an external power supply, just like the graphics industry legend, the four-processor Voodoo 5 6000 from 3dfx used to! The external PSU makes the EAX1800 XT TOP stable in computers with rather low-power system PSUs. Well, this may make sense considering that the Radeon X1800 XT consumes over 100 watts of power working at its default frequencies. We’ll check the power requirements of the ASUS EAX1800XT TOP below in the Power Consumption section.
The included power supply is manufactured by Ilan Electronics, a major producer of power supplies for notebooks on Taiwan. It supports input voltages of 100 to 240VAC and yields an output current of 6.67A at 12VDC. Easy to calculate, its wattage is exactly 80 watts. Our sample of the ASUS EAX1800 XT TOP was meant to sell in North America, so the power supply had an appropriate power cord that ended in a plug with two flat and one round contact.
Nothing prevents you from using the EAX1800XT TOP without the additional power supply, though. To do so, detach the cable from the 6-pin connector on the card and replace it with an appropriate connector from your system PSU. If necessary, you can use the included Molex à PCI Express adapter.
Like the ordinary Radeon X1800 XT, the ASUS EAX1800XT TOP uses Samsung’s K4J52324QC-BJ12 GDDR3 memory. Eight such chips, 512Mb each, give you a total of 512 megabytes. Having an access time of 1.25 nanoseconds, the memory is meant to work at 800 (1600) MHz and this is indeed the frequency it is clocked at on the EAX1800XT TOP. The graphics processor frequency is 700MHz, which is 75MHz above the reference card’s core clock rate. We’ve only once met so high a core frequency before; it was the much simpler Chrome S27 graphics processor from S3 Graphics (for details see our article called S3 Chrome S27 Graphics Processor Review: Worthy Performance for Its Class?).
Among other interesting features of the EAX1800XT TOP we should note the larger heatsink on the power circuit’s transistors and the Rage Theater chip that endows the card with VIVO functionality.
The EN7800GTX TOP graphics card was equipped with a slightly improved version of the Arctic Cooling NV Silencer 5 cooler. Here, ASUS has put its own cooler on the EAX1800XT TOP, although Arctic Cooling has something to do with it, too. At least the fan is actually made by Arctic Cooling.
This time the cooler incorporates heat pipes. These pipes take heat away from the copper sole that contacts directly with the GPU die and distribute it uniformly in the heatsink made up of thin aluminum ribs. The use of heat pipes helped make the copper foundation thinner and thus minimize the overall weight of the cooler.
The airflow created by the blower goes through the heatsink and to the back of the PCB, also cooling the heatsink on the MOSFETs. The air-directing casing is made of black plastic and has a mirror insert with an ASUS logo that shimmers with all the colors of the rainbow. Thick dark-gray thermal paste is placed between the heatsink’s sole and the GPU whereas the memory chips make contact with the cooler sole via thick elastic pads of some rubber-like material.
This cooling system resembles the Accelero X2 cooler in design, which points at its developer, Arctic Cooling, who created this solution for ASUS. We don’t know if ceramic or fluid bearings are used in the fan, but the main advantage of Arctic Cooling fans has always been their very low level of noise, so we can expect the ASUS EAX1800XT TOP to be very quiet, too.
We were very curious to see how much power was needed for an ASUS EAX1800XT TOP, now that the card comes with an external power supply. How much more than the standard Radeon X1800 XT does it consume? The power consumption measurements were performed on a special testbed with the following configuration:
We measured the power consumption of the cards with a digital multimeter Velleman DVM850BL (0.5% measurement accuracy). To put a Peak 3D load on the card we ran the first SM 3.0 graphics test from 3DMark06 in a loop at 1600x1200 resolution and with enabled 4x full-screen antialiasing and 16x anisotropic filtering. Then we created an extremely high 2D load by launching the 2D Transparent Windows test from Futuremark PCMark05. Idle Mode means that the graphics card was just showing the Windows XP Desktop and the computer wasn’t running any resource-consuming applications. Here are the results:
Against our expectations, we don’t anything extraordinary here. The graphics card from ASUS consumes about as much power as the newer Radeon X1900 XT. The latter has 48 pixel processors, but works at lower frequencies. The power consumption grew a mere 7W above that of the Radeon X1800 XT, which is not much at all.
There is obviously no need for the external power supply if you’ve got a good-quality 450W PSU in your system. But if your system PSU cannot stand a high load on the +12V line for long or if its wattage is below 400W, the external power supply may save you the trouble of upgrading your PSU to use the EAX1800XT TOP graphics card.
We don’t think this is a likely scenario, however, because high-end computers that high-performance graphics cards like the ASUS EAX1800XT TOP are bought for surely have a high-wattage and reliable PSU.
As we had anticipated, the cooling system of the ASUS EAX1800XT TOP proved to be near silent thanks to the Arctic Cooling blower that works on quiet bearings and features an optimal design of the blades. We haven’t seen too many silent graphics cards in our tests, but this one is among them. You shouldn’t worry at all that the EAX1800XT TOP may disturb your quiet.
Dealing with a pre-overclocked product, we hadn’t expected the EAX1800XT TOP to be any good at overclocking. And that was true: the card was only stable at 730/1720MHz. Such a small frequency growth cannot affect the card’s performance much, so we didn’t test it at overclocked frequencies. This is certainly not an overclocker-targeted product after all.
Like the overwhelming majority of today’s graphics cards that follow the reference design and belong to a high enough class, the ASUS EAX1800XT TOP delivered an excellent 2D image quality in all display modes available with our laboratory monitors, up to 1800x1440@75Hz. We couldn’t see any undesired artifacts like fuzzy text or shadowing on the screen.
To test ASUS EAX1800XT TOP we assembled the following test platform:
The ATI and NVIDIA drivers were set up as usual.
ATI Catalyst :
Nvidia ForceWare:
We selected the highest graphics quality settings in each game, the same for ATI’s and Nvidia’s solutions, except for the Pacific Fighters flight simulator which requires vertex texturing support to enable its Shader Model 3.0 mode. The Radeon X1000 family doesn’t support this feature and runs the game in the Shader Model 2.0 mode. We did not edit the games’ configuration files, we used only the configuration options offered by the gaming engine. If possible, we used the games’ built-in benchmarking tools and if not, we measured the frame rate with the FRAPS utility. We measured minimal as well as average fps rates possible.
If the games allowed adjusting the full-screen anti-aliasing and anisotropic filtering settings, we enabled FSAA 4x + AF 16x. Otherwise, we forced the necessary mode from the graphics card driver. If any of the gaming engines didn’t support FSAA, we didn’t run the tests in eye candy mode. Since the overclocking potential of the ASUS EAX 1800XT TOP turned out almost null, we decided not to run any tests of the overclocked platform, because the slight increase in the GPU and memory frequencies was very unlikely to affect the performance of the card in most tests. Besides ASUS EAX1800XT TOP, we also tested the following graphics solutions:
These games and applications were used as performance benchmarks:
First-Person 3D Shooters
Third-Person 3D Shooters
Simulators
Strategies
Semi-synthetic benchmarks
Synthetic benchmarks
The higher GPU and memory clock rates ensure the ASUS EAX1800XT TOP a certain advantage over the standard Radeon X1800 XT in Battlefield 2 . Yet it is only in high resolutions that the ASUS nearly overtakes the Radeon X1900 XT – all thanks to its faster graphics memory. Nvidia’s GeForce 7800 GTX 512 is still in the lead at the “pure speed” settings as well as in the first two resolutions of the “eye candy” mode.
The ASUS EAX1800XT TOP is only 3-5fps better than the Radeon X1800 XT whereas the Radeon X1900 XT is noticeably faster, by up to 40%, and runs The Chronicles of Riddick at a comfortable speed even in 1280x1024 with turned-on FSAA and anisotropic filtering. Overclocking the Radeon X1800 XT is of almost no effect here while the 48 pixel processors and the 50% larger HyperZ buffer help get closer to the results of the GeForce 7800 GTX 512 despite the fact that the latter has a more efficient OpenGL driver.
The increase of the GPU clock rate from 625MHz to 700MHz brings certain dividends in Call of Duty 2 . The ASUS card manages to outpace the Radeon X1900 XT two times here, and one time by as much as 20%! Yes, the Radeon X1900 XT carries as many as 48 pixel processors on board and supports the Fetch4 feature for a faster rendering of dynamic shadows using the shadow maps technique, but the boosted frequencies of the ASUS EAX1800XT TOP easily make up for all that.
Here, it is the higher speed of the raster operators (ROPs) that seems to contribute a lot to the overall result. The memory frequency, on the contrary, doesn’t matter much as is indicated by the negligible difference between the Radeon X1900 XT, Radeon X1800 XT and ASUS EAX1800XT TOP in the hardest test mode (1600x1200, 4x FSAA + 16x AF). By the way, this is the single mode in which the EAX1800XT TOP cannot give you a frame rate of 60fps which is necessary for smooth and comfortable play.
Doom 3 and The Chronicles of Riddick are two very similar applications from the technological point of view. Particularly, they both use OpenGL and stencil shadows, thus handicapping the Radeon X1000 architecture. And just like in The Chronicles of Riddick , there is almost no effect from overclocking the Radeon X1800 XT. The ASUS card can yield just a few extra fps in this game: the EAX1800XT TOP is nowhere ahead of the Radeon X1800 XT by more than 2-3fps. The absolute performance of the card is high, though. You will only find the resolution of 1600x1200 at the “eye candy” settings unplayable. If you do want to play the game at such settings, consider the GeForce 7800 GTX 512 which feels quiet at home under such conditions.
We couldn’t find any differences among the top-end graphics cards on the Pier map. They all have roughly the same speed here even in high resolutions and when full-screen antialiasing is enabled.
The Research map can still serve as a measure of performance of such devices as Radeon X1800, Radeon X1900 or GeForce 7800. The frame rate being about 83-88fps in the hardest test mode, we can state that there is almost no effect from overclocking the Radeon X1800 XT in this game, too, even though it is completely different from OpenGL shooters like Doom 3 and The Chronicles of Riddick . Meanwhile, the Radeon X1900 XT stops very short of the 100fps mark so the architectural innovations implemented by ATI are not so unrewarding.
It can hardly be the 48 pixel processors – there seems to be little work for them in Far Cry . But the larger hierarchical Z-buffer and the additional general-purpose registers may be the factors that list the Radeon X1900 XT up to the same level with Nvidia’s card in this test.
The ASUS EAX1800XT TOP behaves strangely in F.E.A.R. , too. The increased GPU clock rate should have positively affected the speed of pixel shaders this game abounds in. The overall speed of the game should have risen as a consequence, but we can’t see any trace of it. The minor performance gain in high resolutions may as likely as not be due to the increased memory frequency.
Half-Life 2 doesn’t give us too much thinking matter since the participating graphics cards are all limited here by the performance of the central processor our testbed is equipped with. The Athlon 64 4000+ isn’t the most advanced CPU for today and has been acting as a bottleneck in our tests more often lately, so we are going to replace it with an Athlon 64 FX-60 in near future. Still we can note that the ASUS EAX1800XT TOP is 1-2fps slower than the Radeon X1800 XT in high resolutions of the “eye candy” mode for some unclear reason.
The Half-Life 2: Lost Coast tech demo is quite a different story, although it produces a rather ambiguous picture of the graphics cards’ performance. In the first two resolutions, the speed of the ASUS card differs but little from that of the Radeon X1900 XT, whether full-screen antialiasing is on or off, but the latter is suddenly ahead in 1600x1200, especially at the “pure speed” mode where the memory subsystem performance is of little importance. The gap between the two cards amounts to 30% in this case! Can it be that the 48 pixel processors of the new chip from ATI are necessary for one resolution only? Maybe it’s the enlarged Z-buffer rather than the pixel processors? It may be so since the resolution of 1600x1200 is the only one where the Radeon X1900 XT outperforms the GeForce 7800 GTX 512 that has an ability to process 32 Z-values per clock cycle. The Radeon X1900’s shadow mapping acceleration feature may have played some role here, too.
The considerable GPU/memory frequency growth doesn’t translate into a higher performance in this game. The ASUS EAX1800XT TOP is hardly any better than the standard Radeon X1800 XT or the Radeon X1900 XT, and is much slower than the GeForce 7800 GTX 512. It may be due to some feature of the game engine or the result of the GeForce 7800 GTX 512 having 24 TMUs, yet niether of these suppositions should give you a cause for worry. The minimum performance of all the top-end Radeon X1000 models is never lower than 60fps in Project: Snowblind.
Something like in the previous test can be observed in Quake 3 , too. The performance of the ASUS EAX1800 TOP is just a little higher than that of the Radeon X1800 XT in the “pure speed” mode. When we switch to the “eye candy” settings, the graphics cards on ATI’s R520/R580 chips just have the same speed irrespective of display resolution. They are a little slower than the GeForce 7800 GTX 512 in the hardest case, yet the frame rate they provide (95-96fps) should be quite enough for any gamer.
With its multiple texture lookups, Serious Sam 2 definitely favors the GeForce 7800 architecture. As for the ASUS EAX1800 XT TOP, its 800 (1600) MHz memory helps it surpass the other Radeons in high resolutions. Moreover, it overtakes the GeForce 7800 GTX 512 in 1600x1200 at the “eye candy” settings – the latter’s memory is 50 (100) MHz faster, but is governed by a less advanced controller than the one the ATI R520 and R580 graphics processors are equipped with.
The speed ceiling goes lower for the graphics cards from ATI Technologies in Unreal Tournament 2004 . There is no difference at all between the Radeon X1900 XT, Radeon X1800 XT and Asus EAX1800XT TOP, each of these cards yielding a speed of over 100fps.
The ASUS EAX1800XT TOP luckily makes up for its fewer pixel processors with high clock rates and is thus quite competitive against the Radeon X1900 XT. Note that the latter ensures a bigger speed reserve in 1600x1200 than the ASUS card does while the average performance of the cards is similar. So, there is some profit from the architectural innovations implemented in the Radeon X1900 in Price of Persia , too.
It is the speed of the pixel processors that is the most important factor in the “pure speed” test mode. This is why the ASUS EAX1800XT TOP outperforms the Radeon X1800 XT by about 10-15%, but it also slower than the Radeon X1900 XT by 8-10%. In this case the 48 pixel processors of the latter card make a difference as the game makes wide use of Shader Model 3.0 and HDR.
The memory subsystem performance then comes to the fore in the modes with full-screen antialiasing and the Radeon X1900 XT finds it impossible to stay ahead of the ASUS card whose memory is clocked at a higher frequency: 800 (1600) MHz against 725 (1450) MHz. These two graphics cards are equals then, providing a minimum frame rate of over 40fps and an average frame rate of 58-61fps in 1600x1200 which is quite enough for a third-person shooter game.
Compare the results of the Radeon X1800 XT and the Radeon X1900 XT and you will see that Colin McRae Rally 2005 is rather indifferent to the number of pixel processors. The larger Z-buffer helps the Radeon X1900 XT win the resolution of 1600x1200, though. On the other hand, the results of the ASUS EAX1800 XT suggest that the game reacts eagerly to any increase of the graphics processor clock rate, so there must be some limitations from the TMUs section, too.
As for the “eye candy” mode, the graphics cards with ATI’s chips are in the lead there, the Radeon X1900 XT making its way to the top with the help of its architectural innovations while the ASUS EAX1800XT TOP, by its increased clock rates. The ASUS is even quite far ahead of the Radeon X1900 XT in 1280x1024: the difference is over 10%.
Being limited by their 16 texture-mapping units, the Radeon X1800 and Radeon X1900 XT can be said to have some troubles with texturing speed as opposed to the GeForce 7800 GTX that has 24 TMUs. The ASUS EAX1800XT TOP can at least partially make up for that deficiency in high resolutions by means of its 700MHz GPU clock rate. But in low resolutions the game engine, originally developed with the GeForce 6/7 architecture in mind, puts down an insurmountable barrier at 65-66fps before the Radeon X1800/X1900 cards. Moreover, the Radeons do not support vertex texturing and thus deliver a lower-quality picture than the GeForce 7800 in this game. The ASUS EAX1800 TOP ensures a comfortable frame rate in resolutions up to 1280x1024 with turned-on full-screen antialiasing and anisotropic filtering, but you should turn to the GeForce 7800 family to enjoy Pacific Fighters in full.
This game’s speed depends more on how fast pixel shaders are executed, yet the texturing speed is important, too. The game’s shaders seem to have multiple texture lookups, just like in Serious Sam . This is why the GeForce 7800 GTX 512 is on top and the Radeon X1900 XT finds itself behind the modest GeForce 7800 GTX in low resolutions. The ASUS EAX1800 XT TOP does well in comparison with the standard card, especially in the “eye candy” mode, but it is only in 1024x768 that it can reach the level of the Radeon X1900 XT.
It is like what we saw in Pacific Fighters : all the R520/R580-based cards stop at a barrier they cannot overcome irrespective of the resolution and our use of FSAA or anisotropic filtering. This doesn’t prevent the ASUS EAX1800XT TOP from offering a rather high average performance in every mode, although the minimum frame rate is rather low, below 25fps even, at the “eye candy” settings.
Aquamark3 isn’t the best benchmark to check the performance of modern top-end graphics cards with. They all show almost the same results here. The GeForce 7800 GTX 512 is the only one to break away from the common group. The ASUS EAX1800XT TOP isn’t any worse or better than the Radeon X1800 XT or the Radeon X1900 XT.
It’s like in the previous test, but the higher GPU clock rate improves the texturing speed of the ASUS EAX1800XT TOP and helps it notch a score of 6000 points, leaving the Radeon X1900 XT and Radeon X1800 XT behind by 90 and 94 points, respectively.
There is almost nothing for the Radeon X1900 to exercise its talents upon in 3DMark03, except that its enlarged HyperZ buffer may be useful at stencil shadows rendering. On the other hand, the frequency growth of the ASUS EAX1800 XT TOP may come in handy as it makes up somewhat for the GeForce 7800 GTX 512’s having more TMUs (24 against 16). The ASUS card cannot beat the latter, of course, but it manages to leave the rest of the test participants behind anyway.
The ASUS EAX1800XT TOP is just a little faster than the Radeon X1800 XT and Radeon X1900 XT. Its advantage is not bigger than 10% in any of the test modes. And this is not enough to overtake the GeForce 7800 GTX in resolutions above 1024x768 at the “pure speed” settings.
In the second test the ASUS card boasts the highest results among all the participating devices on ATI Technologies’ chips even though the Radeon X1900 XT is equipped with a larger Z-buffer. Moreover, the 800 (1600) MHz memory helps the EAX1800XT TOP outperform the GeForce 7800 GTX 512 in high resolutions of the “eye candy” mode.
There’s a similar picture in the third test, but the ASUS EAX1800XT TOP can only overtake the GeForce 7800 GTX 512 in the highest resolution of the FSAA + AF mode.
The ASUS card keeps up on the same level with the Radeon X1900 XT in the fourth test, with minor variations into either side depending on the test mode and resolution. The ASUS enjoys a 10-12% lead over the Radeon X1800 XT that is clocked at the standard frequencies.
So the overall score of the ASUS card is confirmed by the results of the separate 3DMark03 tests.
3DMark05 has quite a different story to tell. It uses complex pixel shaders, so it is no wonder the overclocked Radeon X1800 XT (we mean the ASUS EAX1800XT TOP, of course) finds it impossible to compete with the 48 pixel processors of the Radeon X1900 XT. Anyway, the ASUS card delivers the performance of a GeForce 7800 GTX 512 in this benchmark, which is a superb achievement. Let’s examine the results in more detail now.
In the first test the ASUS is less than 10% ahead of the ordinary, non-overclocked Radeon X1800 XT and is also 10-20% behind the Radeon X1900 XT. The gap between the ASUS EAX1800XT TOP and the GeForce 7800 GTX 512 is smaller, especially in the “pure speed” mode.
The memory subsystem performance is not that important in the second 3DMark05 test, so the ASUS EAX1800XT TOP manages to keep up the tempo of the GeForce 7800 GTX 512 even in the most resource-consuming modes. The Radeon X1900 XT still has the best results among the graphics cards included in this test session.
It’s all generally the same in the third test, although the GeForce 7800 GTX 512 enjoys a certain advantage. Although this test uses textures-heavy shaders, the Radeon X1900 XT beats the Radeon X1800 XT as well as the GeForce 7800 GTX 512; the GeForce is getting somewhat closer to the leader in the “eye candy” mode.
The results of the ASUS EAX1800XT TOP in the individual 3DMark05 tests quite agree with the overall result: this solution is a little faster than the Radeon X1800 XT, but slower than the Radeon X1900 XT. The ASUS is about as fast as the GeForce 7800 GTX 512 in this benchmark.
As opposed to 3DMark05, overclocking the Radeon X1800 XT brings about a bigger effect in 3DMark06 and helps the ASUS card score over 4000 points, although it still cannot rival such solutions as Radeon X1900 XT or GeForce 7800 GTX 512.
The ASUS EAX1800XT TOP graphics card is closer to the leaders in the SM2.0 tests where it enjoys the biggest boost from overclocking and finds itself just a very little behind the Radeon X1900 XT and GeForce 7800 GTX 512. This seems to be the result of the benchmark’s using rather simple shaders that fit within the Shader Model 2.0 framework. Moreover, the second SM2.0 graphics test depends greatly on the vertex processor performance, and this may have affected the overall result, too.
Quite expectably it is the Radeon X1900 XT that wins the SM3.0/HDR tests – the power of its 48 pixel processors is not to be underestimated. They are called for in this test, especially in the airship scene.
Despite its higher memory frequency, the ASUS EAX1800 XT TOP falls farther behind the Radeon X1900 XT as the display resolution grows up. The smaller Z-buffer and the lack of Fetch4 support drag the ASUS card down here. On the other hand, the GeForce 7800 GTX 512 doesn’t need any of these features as it has 24 TMUs and high-performance memory instead; it is just a little behind the Radeon X1900 XT in high resolutions and is in the lead in low ones.
The Radeon X1900 XT is not so far ahead of the ASUS EAX1800XT TOP in the second graphics test whereas the GeForce 7800 GTX 512 is not such a confident leader in high resolutions anymore. We can rather say it is roughly as fast as the two solutions on ATI’s chips there.
You could have expected something like that from the first SM3.0/HDR graphics test: the 48 pixel processors of the Radeon X1900 XT help it break away from the GeForce 7800 GTX 512 which has half the number of pixel processors and clocks them at a considerably lower frequency, too. Meanwhile, the ASUS EAX1800XT TOP is 20% faster than the standard Radeon X1800 XT and is thus capable of competing with the Nvidia card even though the shaders in this test contain numerous texture lookups.
The “eye candy” mode is interesting because the gap between the Radeon X1900 XT and the ASUS EAX1800XT TOP is increasing towards higher resolutions, and the faster memory of the latter card cannot do anything about that. The larger Z-buffer of the Radeon X1900 must again be the reason for such results.
The second SM3.0/HDR graphics test is much simpler than the first one and doesn’t require the graphics cards to do a lot of math1, so the Radeon X1900 XT cannot beat the GeForce 7800 GTX 512 as effectively as it did in the first test.
The test focuses on HDR effects and realistic dynamic shadows. The Radeon X1900 XT has a considerable advantage over the ASUS EAX1800 XT TOP – we mean the Fetch4 feature which helps taking four samples from single-component textures; and shadows maps are single-component textures. Despite its high operating frequencies of 700MHz GPU and 800 (1600) MHz memory, the ASUS card cannot do any better than the GeForce 7800 GTX.
You can see a curious thing happening when we turn on full-screen antialiasing and anisotropic filtering. The positive effect from the increased frequencies of the ASUS EAX1800XT TOP diminishes as the display resolution grows up, and this graphics card eventually become as fast as the ordinary Radeon X1800 XT. There must be some additional limiting factor; perhaps the card’s Z-buffer is not large enough.
So it is all quite clear in 3DMark06: the ASUS EAX1800XT TOP is not much slower than the Radeon X1900 XT in the SM2.0 tests, but in the SM3.0/HDR tests that use numerous and more complex shaders the gap between these two graphics cards is bigger. Do not forget that the Radeon X1900 XT is more efficient at processing shadows created by means of shadow maps than the ASUS card which is based on the R520 GPU that doesn’t support the Fetch4 feature.
The ASUS EAX1800XT TOP/2DHTV graphics card has left us highly pleased, as many other cards from ASUS we’ve tested in our labs have done. Above all we liked its near silent yet very efficient cooling system. Its single drawback in comparison with the older Arctic Cooling systems is that it does not exhaust the hot air outside the PC case and thus doesn’t help improve the thermal conditions of the other system components. There shouldn’t be any trouble if your system case is ventilated well; otherwise you may have some overheat-related problems because the ATI R520 chip overclocked to 700MHz generates quite an amount of heat.
What we don’t quite understand is the purpose of the external power supply the EAX1800XT TOP comes with. This graphics card consumes a mere 7 watts more than the regular Radeon X1800 XT, and enthusiasts who buy and use such top-end graphics cards surely have high-wattage system PSUs in their computers already. On the other hand, why not? You don’t have to use the external power supply if you don’t want to, and if your system PSU isn’t powerful enough, you won’t have to buy a new one. So after all, it is not a bad thing to receive an additional power supply with the card.
The gorgeous set of accessories that the ASUS EAX1800XT TOP comes with must also be mentioned. Besides the graphics cards proper, you receive a good 10-button gamepad with feedback, very suitable for playing arcade shooters like Peter’s Jackson: King Kong, Prince of Persia or Splinter Cell: Chaos Theory. You’ll also find video-editing software in the pack, although the old version of the Rage Theater chip isn’t too good at capturing video. It will do well enough, however, if you are going to convert your VHS archive into digital form.
As for performance, we do testify to ASUS’s claim that its pre-overclocked version of the Radeon X1800 XT is 10% faster than the standard version of the card. In some cases – when the performance is limited by the number of TMUs – the EAX1800XT TOP is even faster at times than the Radeon X1900 XT. And still, though the ASUS card rightly deserves the title of the best Radeon X1800 XT, it cannot claim the name of a GeForce 7800 GTX 512 killer notwithstanding the solid GPU and memory frequency boost – it is a prerogative of the ATI Radeon X1900 family.
This is why the value of the ASUS EAX1800XT TOP/2DHTV for the potential customer is unclear. The ASUS EAX1900XT/2HDTV provides a better performance and offers a future-oriented reserve of 48 pixel processors. The rich accessories and the noiseless cooling system may be arguments in favor of the EAX1800XT TOP, yet we’d advise you to wait a little – ASUS will surely release the TOP version of the EAX1900XT/XTX one of these days.
Highs:
Lows: