Report : 1, 2 or 3 graphic cards? - BeHardware
>> Graphic cards
Written by Damien Triolet
Published on February 8, 2008
URL: http://www.behardware.com/art/lire/699/
Page 1
Introduction
After having reintroduced the possibility of using two graphic cards in parallel in order to increase performances with SLI, and Quad SLI’s big failure which tried to use two double graphic cards, Nvidia decided at the end of 2007 to add 3-way SLI to its multi-GPU offers. This gives us the occasion to take full inventory on the performances of SLI and CrossFire configurations.
 The multi-GPU system The nature of 3D rendering means that it’s very parallelizable, in other words, the same operation is executed on a significant number of distinct elements and on several levels. Geometry processing and that of its numerous vertices/triangles can be done in parallel as well as the calculation of pixels, and finally of course, the processing of images.

It’s therefore obvious that here lies an opportunity to divide up the required tasks of 3D rendering so that several graphic processors can be used. The issue then becomes one of finding the most efficient combination out of all of these possibilities and implementing it in a robust and performance consciousness way to drivers.
Sharing the work in terms of geometry between two processors inside the same image is feasible but rather complex to put into place. Gains may not be too interesting, all the more so that geometry processing represents a very simple task in many games.
Then there is separating geometry processing from the calculation of pixels with one graphic processor exclusively dedicated to one of these tasks and a second one to the other. Unified architectures actually enable GPUs to devote all of their calculation capacities to a very specific task; however, this goes against the very principle of these unified architectures which is the sharing of their processing capacities.
Dividing up the calculation of pixels between several GPUs is a better adapted solution at least with classic GPUs which have processing units dedicated to the calculation of vertices and pixels. In this case, each GPU will entirely process all vertices (in some ways a waste of resources, but either way this would mean dedicated units with nothing to do if this wasn’t the case) and only some pixels, for example, those on half of the screen, one out of two lines, or small zones on the image. Here, we are speaking of SFR (split frame rendering).
The division of work on pixels can be static (fixed once and for all) or dynamic (depending on the complexity of calculation of each zone of the screen and the power of each GPU). If the division is dynamic, this solution has the advantage of being able to efficiently divide the work between two GPUs of differing power; however, significant performances gains can be difficult to achieve in certain situations. While its basic functioning is "simple", it can become extremely complex with the engines in more advanced games.
Currently, the most efficient solution is to share the calculation of images between GPUs. With AFR (alternate frame rendering), each GPU handles the complete rendering of an image not having to worry about the other GPUs. The only challenge is that there be no dependence between images. While it is rather simple to avoid this, in a number of cases, the responsibility falls on the game developer and not the GPU manufacturer who is more or less powerless on this level. It was for this reason that the previous method was initially favored before it was progressively replaced in the majority of cases by higher performance AFR.
Dynamically sharing threads, or diverse operations that are to be executed on various elements, is the next step. More specifically, this will involve sharing the required calculations between different chips in a similar way to what happens between different units on the same chip. There are still a number of obstacles to be overcome, notably in terms of communication between GPUs and their memory, but it seems evident this will be the path to follow in the future.
Page 2
The Quad SLI fiasco, is 3 better than 4 ?The Quad SLI fiasco The arrival of triple SLI may bring back bad memories for some. Launched in partnership with Dell at the start of 2006, Quad SLI was made available to more partners in the following months, and finally there was free access at the beginning of summer of the same year with the arrival of the GeForce 7950 GX2. However, from the start this technology created a few concerns, because besides a few benchmarks, it wasn’t functional. In the beginning, we were told that drivers were too new and then later that Windows XP was at fault.
In fact, Quad SLI never really functioned. Its unfortunate owners were wonderfully misled and Nvidia’s justifications on the subject weren’t too convincing. In our opinion, the reason for this fiasco is that the marketing department decided to launch the technology too early in order to draw a maximum of interest; however, it was before having real visibility on the amount of work necessary to develop drivers that actually live up to this name. Nvidia must have realized that the required resources for their rapid development was too great compared to the small market that Quad SLI represented and they decided against it. The GeForce 8 was about to arrive as was Windows Vista creating an enormous development load which, as you may have understood, had priority.
So to lose time with a Quad SLI that was mainly meant to boost their image by having the current most powerful graphic rendering was thus out of the question. However, given that the products were already launched, Nvidia did not react and placed the fault on Windows XP in response to criticism. Even if this might have been true, feasibility should have been addressed before. To sell a technology, whatever it is, before it is functional is still unacceptable. And for this reason, the same fears returned with the arrival of 3-way SLI.Is 3 better than 4 ? It’s obvious that before launching 3-way SLI, Nvidia wanted to be sure to not fall in the same trap, all the more so that all eyes were on this manufacturer. So why should it work better now than 2 years ago? For several reasons.
A commonly mentioned one is that Windows Vista facilitates the division of images between more than two GPUs. This is probably the case but to a much smaller extent than what Nvidia would have us believe. For them this is a good pretext partly to rid themselves of the blame for the failure of Quad SLI and also for a drastic reduction in the investment in drivers destined for Windows XP. So for the moment, 3-way SLI is exclusive to Windows Vista.
In our opinion, the main reasons that 3-way SLI and even a "new" Quad SLI are more viable today can be summed up in 2 points. First of all, there are now more mature drivers which represent a real investment in terms of multi-GPU rendering as this is the way of the future. Second, games are taking into account multi-GPU systems more and more.
It therefore becomes increasingly possible to "simply" use AFR rendering in which the management of 2, 3 or even 4 GPUs is more or less the same. We are thus far from quad-SFR or an AFR of SFR which were the modes used with Quad SLI in 2006, or rather those that were announced but which Nvidia never really mastered.
Page 3
The systems, the specificationsThe various systems To test the different multi-card solutions, we should speak of systems because not all are compatible with all motherboards. Thus, Nvidia’s SLI is exclusively reserved to its own chipsets (except with Intel’s Skulltrail platform). As for triple-SLI, it’s currently limited to the nForce 680i and 780i. On AMD’s side, there is more flexibility because CrossFire is functional with, in addition to its own chipsets, the following ones from Intel : P35, P965, 975X and X38.
 Our CrossFire platform was based on the X38 which we preferred to an AMD chipset because the Phenom platform isn’t too convincing yet. Making the transition to a Core 2 and still keeping CrossFire, an Intel chipset had to be used. We opted for an Asus P5E motherboard which allowed us to stay in DDR2 like with Nvidia platforms. Two Radeon HD 3870s were combined for graphics.
 For the SLI platform, we used an eVGA nForce 680i motherboard. It is limited to PCI Express 1.1 but as you can see in our review of the nForce 780i, its PCI Express 2.0 support isn’t ideal and doesn’t add any significant gains. Also, the GeForce 8800 Ultra does not support PCI Express 2.0. In addition to the GeForce 8800 Ultras, we also tested a couple of GeForce 8800 GTs.
 Finally, the 3-way SLI platform is of course identical except there are 3 GeForce 8800 Ultras combined to create the ultimate system in terms of current graphic power.
Specifications  While 3-way SLI allows blowing away graphic capabilities, it doesn’t come without a price as you will have to invest in three GeForce 8800 Ultras or GTXs.
You may recall that while calculation power and memory bandwidth are added up in a multi-GPU system, it isn’t the same for the quantity of memory as data is duplicated in each local memory. Also, multi-GPU functioning means use of more memory because a zone on the "master" graphic card has to be set aside whose sole purpose is to store the work of other GPUs. Consequently, the pressure on memory is increased in both SLI and CrossFire. You should therefore try and avoid at all costs combining low memory graphic cards (HD 3850 256 MB and other GeForce 8800 GT 256 MB), even though they may appear to be good deals.
Page 4
Power consumption, set up, the testPower consumption We evaluated the power consumption of the different systems. Measurements were taken at the wall socket, in order to have the total power consumption of the power supply, in this case a Cooler Master Real Power M1000 (1000 watt). We added the results of a system equipped with a single Radeon HD 3870 but based on an nForce 680i platform like all the GeForce cards in order to have an idea of the difference in consumption related to the motherboard and chipset.
 The use of the 55 nanometer process and PowerPlay to reduce consumption enabled the Radeon HD 3870 to be very economical in stand-by. It’s an entirely different situation for the GeForce 8800 Ultras which use quite a bit of energy. A single one requires more power than two GeForce 8800 GTs!
While the GeForce 8800 Ultra measurements in load could be acceptable because it is a very high end system, at rest the amount is more problematic. Each one of its cards uses almost 100 watts or 300 watts in total uselessly wasted by the whole configuration.
Either way, you will have to make sure that you have an adequate power supply to use such systems.Set up Activating SLI or CrossFire in drivers is very simple as you only have to check a box in their specific panel. Both AMD and Nvidia integrate a hardware and software version of their multi-GPU technology. The hardware version uses dedicated components present in GPUs as well as the specific connector to transfer information (or at least the biggest possible part of it) from one card to another. This is the highest performance mode. The software version works with the PCI Express bus to transfer all the data between cards and is therefore a little less efficient.
  CrossFire and 3-way SLI connections. Of course, we used the hardware version in all cases. For CrossFire, two connectors are currently required, one in each direction; however, AMD told us that in the future only one would be necessary. For Nvidia, a single connector is needed for SLI while a special one is needed for triple SLI which allows interconnecting all cards (the 1st to the 2nd, the 2nd to the 3rd and the 3rd to the 1st). It’s something that doesn’t come with nForce 680i motherboards and has to be acquired while NForce 780i motherboards include it with their bundle.
Note that for SLI, Nvidia also authorizes the software version. This isn’t the case for CrossFire with two Radeon HD 3870s or triple SLI which can only be activated once the special connector is in place.
The test For this test, we used ten games, four of which support DirectX 10. Tests were carried out in 1920x1200 and 2560x1600, the resolution of 30" screens.
Anisotropic filtering as well as HDR were activated in all cases when available in the game. Finally, transparency/adaptive antialiasing were activated in multisampling mode.
Of course, all Windows Vista updates relative to performance and multi-GPU systems were installed.
Configuration Intel Core 2 Extreme QX6850 eVGA nForce 680i SLI (GeForce) Asus P5E (Radeon) 2 GB DDR2 Windows Vista Forceware 169.04 (169.28b for Crysis) Catalyst 8.1 hotfix.
Page 5
Enemy Territory : Quake WarsEnemy Territory : Quake Wars
 While Quake Wars is based on the Doom 3 engine, it has undergone some evolution such as megatexturing which facilitates the work of artists; however, there is the additional cost in terms of decoding and access to megatextures. In the end, Quake Wars is a little more resource heavy than Doom 3 or Quake 4. We saved a demo in a sequence versus 4 bots. Given that artificial intelligence was not calculated in the time demo, results were less affected by the CPU than in actual gameplay or at least in this case versus our bot adversaries.
All parameters were set to a maximum in the game including 16x anisotropic filtering. The patch 1.2 was used.
 In this first test, SLI does better than CrossFire, although in this case, triple SLI is CPU limited.
 Once antialiasing is activated, triple SLI takes the lead; however, the GeForce 8800 GT performances plummet in SLI and in 2560x1600 due to a lack of video memory.
Page 6
Half Life 2 Episode 2Half Life 2 Episode 2
 Still based on the Source Engine, Half Life 2 Episode 2 doesn’t really have anything new on the technological level. It simply optimizes and more heavily relies on the engine’s capabilities, making the game more resource heavy than its previous versions. We carry out a demo with all game options set to a maximum including anisotropic filtering which is in 16x.
 In Half Life 2 Episode 2, CrossFire does better than SLI, especially in "low" resolution. Triple SLI has a problem here and its performances are reduced compared to classic SLI.
 With antialiasing, CrossFire loses its advantage while triple SLI still has some problems.
Page 7
S.T.A.L.K.E.R.S.T.A.L.K.E.R.
 We carry out an identical movement and measure the framerate with fraps. The test was done in high quality, complete dynamic lighting, maximum details (anisotropic filtering 16x) and foliage shadows. S.T.A.L.K.E.R. uses an engine based on differed rendering, which is fundamentally incompatible with MSAA and makes the use of antialiasing impossible. A type of filtering of edges carried out with a shader can be activated but results are mixed. The 1.00004 patch is used.
 CrossFire shows very high gains here and SLI does rather well except for the 8800 GTs in 2560x1600 which seem limited by their 512 MB. For triple SLI, the result is not very interesting as performances are lower compared to classic SLI.
Page 8
Rainbow Six : VegasRainbow Six : Vegas
 The first PC game based on the Unreal Engine 3.0, Rainbow Six : Vegas is still a very resource heavy game. We measure performances in the introductory scene. The HDR mode is activated as it is more or less obligatory as without it banding is very noticeable. Shadows are set to “low” because a higher quality in this domain lowers performance too much in certain areas.
 Originally designed for the Xbox 360, this game seems to have a natural affinity for the Radeon HD which has a similar architecture to the game console’s graphic chip. It’s the same for CrossFire which shows an almost perfect gain here. SLI and triple SLI also display very good performances although even in 2560x1600 SLI is already a big advantage.
 The game does not support antialiasing but Nvidia has implemented it to drivers as it has done with Oblivion. This is contrary to AMD who unfortunately didn’t make this effort. In 2560x1600, triple SLI adds additional fluidity which is quite useful. Gains are relatively high on all levels in this game.
Page 9
OblivionOblivion
 We saved a specific movement in order for it to be always identical and the test reproducible. Of course, HDR was activated.
 All multi-card solutions are CPU limited here and a single graphic card is generally enough even if two add a little more comfort in 2560x1600
 Once FSAA 4x is activated, multi-card solutions become useful. Note the good results of the Radeon HD 3870 in CrossFire which is almost equal to the GeForce 8800 Ultra in SLI. Triple SLI isn’t too useful.
Page 10
Colin McRae DIRTColin McRae DIRT
 To test Colin McRae‘s latest opus which is very resource heavy we carried out a well defined sequence in high quality mode. Note that the activation of antialiasing is highly recommended given the way menus are rendered and because post processing effects amplify aliasing. The patch 1.2 was used.
 Colin McRae is the nemesis of multi-GPU systems. Unfortunately, this reminds us that games that are less flattering to the eye and not often used to demonstrate the latest performance gains do not always receive satisfactory support from AMD and Nvidia. In CrossFire, road side spectators are constantly appearing and disaprearing, and in SLI although perfor mances increase there are huge cuts in fluidity that make the game unplayable.
 It’s the same problem with antialiasing and you will have to deactivate SLI or CrossFire to play this game.
Page 11
BioshockBioshock
 The first game based on the Unreal Engine 3.0 to support DirectX 10, Bioshock has great graphics even in DirectX 9 mode while it is less resource heavy than Rainbow Six : Vegas. We carry out a well defined sequence of movement with all options pushed to a maximum.
 The Unreal Engine 3.0 is definitely very appreciated by multi-GPU systems and gains are very high. Given the extremely high numbers, triple SLI isn’t really of practical interest here.
 Like with Rainbow Six : Vegas, Nvidia allows the activation of antialiasing for this game which doesn’t normally support it while AMD doesn’t offer this option.
 In DirectX 10 mode, GeForce 8 performances are similar to what we obtained in DirectX 9 mode while gains in SLI and triple SLI are more tempered. This isn’t the case for CrossFire but Radeon HD 3870 performances are lower in DirectX 10.
Note that it is not yet possible to activate antialiasing in DirectX 10 mode for Nvidia and AMD.
Page 12
Company of HeroesCompany of Heroes
 Given that Company of Heroes received a DirectX 10 patch that adds a real plus on the graphics level, we decided to add it to our test protocol. All graphic settings were pushed to a maximum except for terrain details which remained on High (Ultra mode is reserved for DirectX 10). Textures were also limited to High, because the game indicates a lack of system memory for the GeForce in DirectX 9 mode with Ultra textures.
We run the integrated test on version 1.71.

Company of Heroes also shows very high gains in multi-GPU mode; however, triple SLI isn't very useful here. Actually, a single card is generally enough.
 It’s the identical situation with antialiasing 4x activated except for demanding gamers that want to play in 2560x1600.
 In DirectX 10, performances are clearly lower as the graphics engine is more advanced. This time two graphic cards are of use even in 1920x1200. In these conditions the Radeon HD 3870 does not allow playing in 2560x1600 with satisfactory quality.
 Triple SLI becomes useful here in 2560x1600.
Page 13
World in ConflictWorld in Conflict
 Very resource heavy and with nice graphics, it’s only natural World in Conflict joins our test suite. We carry out the internal test with the patch 1.0002. All game options are pushed to a maximum. Catalyst 7.12 was used as the 7.11b version wasn’t correctly functioning in CrossFire.
 CrossFire suffers from a rather bothersome bug in DirectX 9 even with Catalyst 7.12 and 8.1. Text was scrambled which made navigation impossible in menus. While SLI is of interest here, this really isn’t the case for triple SLI given the CPU limitation. Given that there is a slightly higher load on this level, its performances are even a bit reduced.
 With antialiasing 4x, GeForce 8800 GT performances in 2560x1600 plummet due to a lack of video memory.
 In DirectX 10, CrossFire functions correctly and even adds a higher gain than SLI which places the Radeon HD 3870s ahead of the GeForce 8800 GTs. Triple SLI still doesn’t stand out.
 GeForce 8800 GT performances also fall here in 2560x1600 and in even higher proportions. Triple SLI shows a gain but it’s too small to be of real use.
Page 14
CrysisCrysis
 An absolute must in terms of gaming, Crysis was tested with its 1.1 patch (optimized for multi-GPU systems) as well as the latest drivers specific to this patch, Forceware 169.28b and the Catalyst 8.1 hotfix. We carried out the internal test in DirectX 9 High mode (forced via the addition of ‘’–dx9’’ behind the executable) and in DirectX 10 Very High. Given the fact that we no longer had access to a 30" screen when the patch came out, we were limited to a resolution of 1920x1200. However, looking at results you can easily see that there is still no possibility of being able to play in 2560x1600 with this game!
 We first have to point out that the GPU demo integrated to Crysis is limited by the CPU to more or less 35 fps. Triple SLI therefore cannot make its presence felt here and even displays a slight reduction compared to SLI given its higher costs in system resources.
 With antialiasing 4x, we have the same situation although the fps is slightly lower. It is therefore possible that triple SLI isn’t really functional in Crysis in DirectX 9 mode. GeForce 8800 GT performances plummet in SLI because it’s a little too much for GPUs with "only" 512 MB.
 In DirectX 10 and with all details pushed to a maximum, triple SLI finally showed a gain though it was small. CrossFire is less efficient here than SLI but this should improve because the Radeon HD 3870 X2 showed higher gains.
 Finally, in DirectX 10 and with antialiasing 4x none of the solutions were up to the challenge. Here again, the GeForce 8800 GT in SLI suffers from a lack of memory.
Page 15
Recap of performancesRecap Although individual game results are interesting, we calculated a performance index based on all tests with the same weight given to each game. A score of 100 was attributed to a GeForce 8800 GT in 1920x1200. You may recall that Crysis was only tested in 1920x1200 and therefore only affects the average results in this resolution.
 On average, the GeForce 8800 GT and Radeon HD 3870 display very close performances; however, in SLI the 8800 GTs have the advantage. Triple SLI has problems setting itself apart because it is often limited by the CPU even in 2560x1600.
 With antialiasing 4x filtering activated, it’s the same situation for triple SLI. While the GeForce 8800 GTs in SLI have very good results in 1920x1200, performances plummet in 2560x1600 in numerous games which places them on average behind a single GeForce 8800 Ultra. As for the Radeon HD 3870, bugged ROPs forced AMD to find an alternative solution that while very flexible limits performances. For this reason, this card does not stand out under these conditions.
Note that results obtained in Rainbow Six: Vegas and Bioshock are not taken into account because the Radeon HD 3870 isn’t capable of supporting antialiasing in these games. Otherwise, you can consult a graph that includes this information in the final grade here.
 With DirectX 10 games, triple SLI adds an advantage especially in 2560x1600. Actually, this isn’t related to DirectX 10 but rather to more complex rendering. SLI and CrossFire also do quite well here.
 In DirectX 10 and with FSAA 4x, it’s very similar except in 2560x1600 where the GeForce 8800 GTs in SLI (and even a single card) show very low performances. With only 512 MB, they are not at ease while the Radeon HD 3870s handle the situation rather well.
Page 16
Gains in 1920x1200Gains in 1920x1200 We put the performance gains from tests into graph form so that they could be interpreted more clearly. If there are no results, this means that the multi-GPU system had a bug that made the game unplayable. A gain of 0% indicates either that there is no gain or that it reduced performances and in this case would be deactivated by the gamer. Therefore there are no negative results. Note that triple SLI gains should be compared to those of GeForce 8800 Ultra SLI and not 8800 GT SLI because they are less limited by the CPU and have more room to express their potential.
 8800 GT SLI and CrossFire gains are generally similar and each has a small advantage depending on the game. Gains were lower with the GeForce 8800 Ultras given that they were more often CPU limited. Triple SLI was of little interest here.
 With antialiasing 4x activated, the load on the graphic system increases and this in theory allows all the solutions to better express their potential. All showed higher gains except the Radeon HD 3870s. For some unknown reason, the poor performances with this filter seem amplified in CrossFire. Finally, triple SLI is likewise of little interest here.
 With DirectX 10 game, triple SLI makes its presence felt while CrossFire is behind due to a reduced gain in Crysis.
 CrossFire bounces back in DirectX 10 and with FSAA 4x showing rather consistent and large gains in the three games tested.
Page 17
Gains in 2560x1600Gains in 2560x1600 We put the performance gains from tests into graph form so that they could be interpreted more clearly. If there are no results, this means that the multi-GPU system had a bug that made the game unplayable. A gain of 0% indicates either that there is no gain or that it reduced performances and in this case would be deactivated by the gamer. Therefore there are no negative results. Note that the gains of triple SLI should be compared to those of GeForce 8800 Ultra SLI and not 8800 GT SLI because they are less limited by the CPU and have more room to express their potential. 2560x1600 should show the most significant gains:
 Indeed, gains are higher here than they were in 1920x1200 which is first and foremost to the advantage of the GeForce 8800 Ultra in SLI and triple SLI.
 With antialiasing 4x activated, CrossFire again shows smaller improvements while without this filter we obtained the opposite result. The GeForce 8800 GTs in SLI also show reduced gains but in their case the problem comes from insufficient memory. For some unknown reason, there is more of an effect on memory for Nvidia than with AMD.
 With DirectX 10 games and in this resolution triple SLI shows an obvious advantage.
 The GeForce 8800 GTs’ lack of memory is once again very evident.
Page 18
Recap of gainsRecap of gains To finish up with all of these numbers, we created a graph that takes all average gains per rendering mode:
 As you can see, overall results do not really set SLI apart from CrossFire. These two technologies are equivalent; however, depending on the product and conditions they alternately take the lead. Besides recent games, you will have to play in 2560x1600 to benefit from triple SLI but this was only to be expected. In theory, gains displayed by the GeForce 8800 GT should always be higher than those from two GeForce 8800 Ultras; however this was not the case as soon as there were increases in complexity. The GeForce 8800 GT’s 512 MB was not always enough.
Page 19
ConclusionConclusion As we often say, we find that multi-card systems are only of interest on the high end in the way that performances can be improved compared to a single card. This is mainly because the rapid evolution of these cards makes an update of your system with a second mid or entry level solution of little interest compared to its replacement.
 This is all the more so that multi-card systems come with a certain number of inconveniences such as higher power consumption, more bulk as well as variable results (sometimes very good, sometimes not) and more bugs than with a single graphic card. Multi-card systems are therefore not yet ready for widespread use. Instead they should be reserved for advanced users that are knowledgeable of the negative aspects and will make sure to always have the latest drivers.
While drivers have strongly evolved as much for Nvidia as AMD (in our opinion SLI and CrossFire being equivalent in this area), they still do not have the same robustness as those with single graphic cards. Little bugs do persist, especially in games that are less in the spotlight.
In the majority of resource heavy games, results are up to expectations. A solution based on two Radeon HD 3870s or GeForce 8800 GTs therefore enables better performances; however this isn’t systematic. The Radeon HD 3870s has results that are slightly disappointing once antialiasing is activated. Also it has performances equal to that of the Radeon HD 3870 X2, which is sometimes more interesting.
For the GeForce 8800 GTs, we also find a limitation in terms of memory in this particular case. For some unknown reason, the GeForce 8 seems either to need more memory than the Radeon HD or else it uses it less efficiently. For this reason, the 512 MB limits performances in certain games in high resolution and with antialiasing activated. Once in SLI, memory use increases again a little more and performances fall instead of improving. This SLI solution is thus recommended for playing in 1920x1200 in complex games, rather than in 2560x1600 with antialiasing and less options or with more simple games.
For those looking for the ultimate performances in high resolutions with antialiasing, they will have to look to the GeForce 8800 GTX or Ultra; however just for the three graphic cards in 3-way SLI this represents a budget as high as 1800 €! It’s an elitist solution that is clearly reserved for wealthier gamers, who have taken the trouble to verify if such a system is adapted to their screen and favorite games because today only a few games will show interesting gains.
Copyright © 1997-2009 BeHardware. All rights reserved.
|