Original Link: http://www.anandtech.com/show/2080



Introduction

Last week saw NVIDIA's introduction of the 7900 GS and the announcement of the 7950 GT. Availability of the 7900 GS has been good since launch, and quite a few manufacturers have released cards above stock clock speeds. We've been able to get ahold of a few of these cards, so we will be able to explore 7900 GS SLI performance. We will also be publishing a 7900 GS mini roundup up in the very near future. For now, here's a list of the cards we have and are currently testing, along with their core/memory clock speeds (though there are more out there).

Albatron GeForce 7900 GS 450/660
EVGA e-GeForce 7900 GS KO 500/690
Leadtek WinFast PX7900GS TDH Extreme 520/700
XFX GeForce 7900 GS 480/700
BFG GeForce 7900 GS OC 540/660

Today also marks the day that ATI slated for the availability of their X1300 XT, X1650 Pro, X1900 XT 256MB, and X1950 series. We currently see very limited availability of the X1950 XTX, X1900 XT 256MB and X1300 XT cards, but the X1950 CrossFire and X1650 Pro are still not to be found. This is certainly one of the major downsides of a paper launch: we don't know when we will see product available. Right now, the products ATI has announced appear very competitive. Their performance looks good, and the ATI stated MSRP is right on target. Unfortunately, both of these aspects could change between now and when we actually see high availability of product on the shelves. Currently, X1950 XTX cards are about $50-$100 higher than expected, as well as being mostly on back order. Some of the X1300 XT cards we've seen are hitting over $100: a price point not even the X1650 Pro is supposed to break. All of this will absolutely factor into our recommendations today.

Aside from all of this, our focus with this second and final part of our NVIDIA GPU refresh series is on the GeForce 7950 GT. We covered the specifications in our previous article, but for a quick and dirty recap, the 7950 GT is essentially an overclocked 7900 GT with 512MB of RAM as opposed to 256MB. The 7950 GT could also be viewed as an underclocked 7900 GTX. All of these cards have the same number of vertex, pixel, and raster pipes, and some 7900 GT cards can even overclock higher than stock 7950 GT speeds. At this level of performance, NVIDIA is targeting a $300 to $350 USD price range. This puts it in competition with the higher priced overclocked 7900 GT cards, as well as the X1900 XT from ATI. To bring it all home, here are our tables of GPU specifications and prices.

NVIDIA Graphics Card Specifications
Vert Pipes
Pixel Pipes
Raster Pipes
Core Clock
Mem Clock
Mem Size (MB)
Mem Bus (bits)
Price
GeForce 7950 GX2
8x2
24x2
16x2
500x2
600x2
512x2
256x2
$600
GeForce 7900 GTX
8
24
16
650
800
512
256
$450
GeForce 7950 GT
8
24
16
550
700
512
256
$300-$350
GeForce 7900 GT
8
24
16
450
660
256
256
$280
GeForce 7900 GS
7
20
16
450
660
256
256
$200-$250
GeForce 7600 GT
5
12
8
560
700
256
128
$160
GeForce 7600 GS
5
12
8
400
400
256
128
$120
GeForce 7300 GT
4
8
2
350
667
128
128
$100
GeForce 7300 GS
3
4
2
550
400
128
64
$65


ATI Graphics Card Specifications
Vert Pipes
Pixel Pipes
Raster Pipes
Core Clock
Mem Clock
Mem Size (MB)
Mem Bus (bits)
Price
Radeon X1950 XTX
8
48
16
650
1000
512
256
$450
Radeon X1900 XTX
8
48
16
650
775
512
256
$375
Radeon X1900 XT
8
48
16
625
725
256/512
256
$280/$350
Radeon X1900 GT
8
36
12
525
600
256
256
$230
Radeon X1650 Pro
5
12
4
600
700
256
128
$99
Radeon X1600 XT
5
12
4
590
690
256
128
$150
Radeon X1600 Pro
5
12
4
500
400
256
128
$100
Radeon X1300 XT
5
12
4
500
400
256
128
$89
Radeon X1300 Pro
2
4
4
450
250
256
128
$79


Before we take a look at performance numbers, we've got a couple retail versions of the 7950 GT in house already. Both EVGA and XFX have sent us cards, and we were quite happy to learn that the XFX card is passively cooled. Up first is a brief look at what we can expect to see from manufacturers on the 7950 GT front.



Retail 7950 GT Cards: EVGA and XFX

In a testament to NVIDIA availability, we have retail products from two different vendors today. Aside from giving us the ability to test SLI with the 7950 GT, we can take a look at what we can expect to see from cards that are currently hitting the shelves. From EVGA, the e-GeForce 7950 GT KO offers overclocked core and memory with a fairly conventional HSF (which covers the RAM in addition to the GPU). XFX has made a very bold move with their 7950 GT and is offering two different silent models: the one we have features an overclock as well. Let's take a look at the EVGA card first.






This card doesn't look much different than we would expect an EVGA card to look. It's a very straightforward design with a copper HSF covering the RAM and GPU. The XFX design is a little more extreme as we can see here.









While we certainly expect overclocked products from EVGA as we have seen many in the past, XFX offering a passively cooled overclocked 7950 GT is quite a statement: the 7950 GT is a very efficient card. In spite of the fact that the X1950 has done a better job on power, heat and noise than the X1900, ATI just doesn't make a card at the $300 price point that could ever hope to be passively cooled. With the 7950 GT essentially being an underclocked 7900 GTX, we are quite impressed that this GPU doesn't need more powerful cooling.



The Test

As promised in part one, we have added SLI tests to the lineup. Most of our setup is the same as the last time, and all single card configurations were tested on the Intel motherboard. To facilitate our SLI tests, we added the ASUS P5NSLI board.

CPU: Intel Core 2 Extreme X6800 (2.93GHz/4MB)
Motherboard: Intel D975XBX (LGA-775)
ASUS P5NSLI
Chipset: Intel 975X
NVIDIA nForce 570 SLI
Chipset Drivers: Intel 7.2.2.1007 (Intel)
NVIDIA nForce 8.22
Hard Disk: Seagate 7200.7 160GB SATA
Memory: Corsair XMS2 DDR2-800 4-4-4-12 (1GB x 2)
Video Card: Various
Video Drivers: ATI Catalyst 6.8
NVIDIA ForceWare 91.47
Desktop Resolution: 1920 x 1440 - 32-bit @ 60Hz
OS: Windows XP Professional SP2


As far as the games go, we have stuck with the same tests this time around. The resolutions we've tested have also shifted a little to focus on the high end. We made sure to hit resolutions that represent the vast majority of LCD panels on the market. Note that while we have not included widescreen performance, the numbers for the common resolutions (1440x900, 1680x1050, 1920x1200) should be within 5-10% of the resolutions we tested. We will be testing without AA for most tests, but 4xAA will be tested in Battlefield 2, Half-Life 2, and Quake 4.



Battlefield 2 Performance

This benchmark is performed using DICE's built in demo playback functionality with a few added extras built in house. When using the built in demo playback features of BF2, frames rendered during the loading screen are counted in the benchmark. In order to get a real idea of performance, we use the instantaneous frametime and frames per second data generated from a benchmark run. We discard the data collected during the loading screen and calculate a result that represents the actual gameplay that was benchmarked. While DICE maintains that results over 100fps aren't reliable, our methods have allowed us to get useful data from high performing systems.

During the benchmark, the camera switches between players and vehicles in order to capture the most action possible. There is a lot of smoke and explosions, so this is a very GPU intensive Battlefield 2 benchmark. The game itself is best experienced with average in-game framerates of 35 and up.

We ran Battlefield 2 using the highest quality graphics settings we could. Shadows, lighting, and especially view distance are very important in order to best play the game. In our opinion view distance should never be set to less than the max, but other settings can be decreased slightly if a little more performance or a higher resolution is required.

Battlefield 2 Performance


At our median resolution of 1600x1200, the 7950 GT splits the difference between the two flavors of X1900 XT. We see just slightly higher numbers for 7950 GT SLI than for the 7950 GX2 setup, while the 7900 GS SLI just leads the single 7900 GTX.



The CPU limit for multi-GPU setups is a little lower than with the single cards, and we do begin to see this at 1280x1024. At the same time, without AA, there really isn't a resolution here that any of these cards will have trouble with. All of our tests fall above 50fps even at the highest resolution.

Battlefield 2 Performance


Enabling AA gives ATI a boost in performance relative to NVIDIA under BF2, so this time we see the 7950 GT losing to the X1900 XT 256MB. 7950 GT SLI increases its margin over the 7950 GX2, while the 7900 GS SLI configuration manages to hang on to the performance spot between singe GPU cards and multi GPU configurations.



CPU limitedness is not a factor with 4xAA turned on. There are three classes of performance that are clearly visible here with the X1900 GT, 7900 GT, and 7900 GS making the lowest performance group. 7950 GT SLI, 7900 GTX SLI, and the 7950 GX2 command the lead, while the rest of the cards fall both well above the lower end cards and well below the higher end multi-GPU configurations.



Black & White 2 Performance

The AnandTech benchmark for Black & White 2 is a FRAPS benchmark. Between the very first tutorial land and the second land there is a pretty well rounded cut scene rendered in-game. This benchmark is indicative of real world performance in Black & White 2. We are able to see many of the commonly rendered objects in action. The most stressful part of the benchmark is a scene where hundreds of soldiers come running over a hill, which really pounds the geometry capabilities of these cards. At launch, ATI cards were severely outmatched when it came to B&W2 performance because of this scene, but two patches applied to the game and quite a few Catalyst revisions later give ATI cards a much needed boost in performance over what we first saw.

A desirable average framerate for Black & White 2 is anything over 20 fps. The game does remain playable down to the 17-19 fps range, but we usually start seeing the occasional annoying hiccup during gameplay here. While this isn't always a problem as far as getting things done and playing the game, any jerkiness in frame rate degrades the overall experience.

We did test with all the options on the highest quality settings under the custom menu. Antialiasing has quite a high performance hit in this game, and is generally not worth it at high resolutions unless the game is running on a super powerhouse of a graphics card. If you're the kind of person who just must have AA enabled, you'll have to settle for a little bit lower resolution than we tend to like on reasonably priced graphics cards. Black & White 2 is almost not worth playing at low resolutions without AA, depth of field, or bloom enabled. At that point, we tend to get image quality that resembles the original Black & White. While various people believe that the original was a better game, no one doubts the superiority of B&W2's amazing graphics.

Black and White 2 Performance


At 1600x1200, the 7950 GT performs very similarly to the 512MB (and more expensive) X1900 XT. While even the 7900 GS is playable here, framerates on the level of the 7950 GT begin to allow the gamer to experiment with AA in the game.



Not much changes with resolution with respect to the 7950 GT. The performance gap between the 7950 GT and X1900 XT does narrow somewhat with increasing resolution, but the two are basically tied in performance. The 7900 GTX SLI becomes CPU limited at 1280x1024, so anything more than 7950 GT SLI is not necessary without a higher resolution or antialiasing. In general, NVIDIA cards seem to scale very slightly better than ATI under Black & White 2 at the tested resolutions and settings.



The Elder Scrolls IV: Oblivion Performance

While it is disappointing that Oblivion doesn't have a built in benchmark, our FRAPS tests have proven to be fairly repeatable and very intensive on every part of a system. While these numbers will reflect real world playability of the game, please remember that our test system uses the fastest processor we could get our hands on. If a purchasing decision is to be made using Oblivion performance alone, please check out our two articles on the CPU and GPU performance of Oblivion. We have used the most graphically intensive benchmark in our suite, but the rest of the platform will make a difference. We can still easily demonstrate which graphics card is best for Oblivion even if our numbers don't translate to what our readers will see on their systems.

Running through the forest towards an Oblivion gate while fireballs fly by our head is a very graphically taxing benchmark. In order to run this benchmark, we have a saved game that we load and run through with FRAPS. To start the benchmark, we hit "q" which just runs forward, and start and stop FRAPS at predetermined points in the run. While not 100% identical each run, our benchmark scores are usually fairly close. We run the benchmark a couple times just to be sure there wasn't a one time hiccup.

As for settings, we tested a few different configurations and decided on this group of options:

Oblivion Performance Settings
Texture Size Large
Tree Fade 100%
Actor Fade 100%
Item Fade 66%
Object Fade 90%
Grass Distance 50%
View Distance 100%
Distant Land On
Distant Buildings On
Distant Trees On
Interior Shadows 95%
Exterior Shadows 85%
Self Shadows On
Shadows on Grass On
Tree Canopy Shadows On
Shadow Filtering High
Specular Distance 100%
HDR Lighting On
Bloom Lighting Off
Water Detail High
Water Reflections On
Water Ripples On
Window Reflections On
Blood Decals High
Anti-aliasing Off


Our goal was to get acceptable performance levels under the current generation of cards at 1600x1200. This was fairly easy with the range of cards we tested here. These settings are amazing and very enjoyable. While more is better in this game, no current computer will give you everything at high res. Only the best multi-GPU solutions and a great CPU are going to give you settings like the ones we have at high resolutions, but who cares about grass distance, right?

While Oblivion is very graphically intensive and is played mostly from a first person perspective (and some third person), this definitely isn't a twitch shooter. Our experience leads us to conclude that 20fps gives a good experience. It's playable a little lower, but watch out for some jerkiness that may pop up. Getting down to 16fps and below is a little too low to be acceptable. The main point to bring home is that you really want as much eye candy as possible. While Oblivion is an immersive and awesome game from a gameplay standpoint, the graphics certainly help draw the gamer in.

The Elder Scrolls IV: Oblivion Performance


The GeForce 7950 GT hits our playable framerate mark at 1280x1024 with a comfortable 23.4 fps. While the X1900 XT does outperform the 7950 GT by a good margin, for the price we still see acceptable performance from NVIDIA's new part.



From the NVIDIA camp, the 7950 GT and 7900 GTX are barely playable at 1600x1200. While the 7900 GS isn't as capable, adding a second card for SLI is a viable option to support this resolution if necessary. Owners of higher end LCD panels will benefit from lower resolutions in spite of the scaling, as the effects are more important in this game. In order to run anything higher than 1600x1200 with these settings, the minimum requirement is a 7950 GT SLI setup. An X1900 / X1950 CrossFire setup is the absolute best option in the case of Oblivion though.



F.E.A.R. Performance

F.E.A.R. has a built in test that we make use of in this performance analysis. This test flies through some action as people shoot each other and things blow up. F.E.A.R. is very heavy on the graphics, and we enable most of the high end settings for our test.

During our testing of F.E.A.R., we noted that the "soft shadows" don't really look soft. They jumped out at us as multiple transparent shadows layered on top of each other and jittered to appear soft. Unfortunately, this costs a lot in performance and not nearly enough shadows are used to make this look realistic. Thus, we disable soft shadows in our test even though it's one of the large performance drains on the system.

Again we tested with anisotropic filtering at 8x, and all options were on their highest quality (with the exception of soft shadows which was disabled). Frame rates for F.E.A.R. can get pretty low for a first person shooter, but the game does a good job of staying playable down to about 25 fps.

F.E.A.R. Performance


All the cards we tested are playable at 1600x1200 under the settings we tested with F.E.A.R. In this case, the 7950 GT performs significantly better than the $200-$250 cards, but the X1900 XT series both have NVIDIA beat at the higher price point. While not necessary at this resolution, multi-GPU configurations scale very well under F.E.A.R.



When looking at multiple resolutions, nothing really changes. Relative performance is very consistent with 1600x1200. Everything remains playable up to 1920x1440, so with most of the tested GPUs enabling 2x/4xAA is definitely an option even at the highest resolutions. The three slowest cards will really struggle with 4xAA at 1920x1440, but the higher end cards along with the multi-GPU configurations should handle that setting.



Half-Life 2: Episode One Performance

Episode One of the new Half-Life 2 series makes use of recent Source engine updates to include Valve's HDR technology. While some people have done HDR that won't allow antialiasing (even on ATI cards), Valve put a high value on building an HDR implementation that everyone can use with whatever settings they want. Consistency of experience is usually not important enough to developers who care about pushing the bleeding edge of technology, so we are very happy to see Valve going down this path.

We use the built-in timedemo feature to benchmark the game. Our timedemo consists of a protracted rocket launcher fight and features much debris and pyrotechnics. The Source engine timedemo feature is more like the nettimedemo of Id's Doom 3 engine, in that it plays back more than just the graphics. In fact, Valve includes some fairly intensive diagnostic tools that will reveal almost everything about every object in a scene. We haven't found a good use for this in the context of reviewing computer hardware, but our options are always open.

The highest visual quality settings possible were used including the "reflect all" setting which is normally not enabled by default, and anisotropic filtering was set at 8x. While the Source engine is notorious for giving great framerates for almost any hardware setup, we find the game isn't as enjoyable if it isn't running at at least 30fps. This is very attainable even at the highest resolution we tested on most cards, and thus our target framerate is a little higher in this game than others.

Half Life 2: Episode 1 Performance


Half-Life 2 Performance is rather uneventful, as all of these cards are completely playable at the very highest settings all the way up to our highest resolution tested. Frame rates will fluctuate throughout the game, but for now at least Half-Life 2 games run without difficulty on all the midrange and faster GPU configurations. It is interesting to note that the 256MB X1900 XT has about a 6% advantage over the 7950 GT throughout our Half-Life 2 testing with AA turned off.

Half Life 2: Episode 1 Performance


Even with 4xAA, everything remains easily playable. Like our sans AA testing, the X1900 XT 256MB outperforms the new 7950 GT across the board. Multi-GPU configurations are not necessary at all under Half-Life 2 without the use of a huge display.



Quake 4 Performance

There has always been a lot of debate in the community surrounding pure timedemo benchmarking. We have opted to stick with the timedemo test rather than the nettimedemo option for benchmarking Quake 4. To be clear, this means our test results focus mostly on the capability of each graphics card to render frames generated by Quake 4. The frame rates we see here don't directly translate into what one would experience during game play.

Additionally, Quake 4 limits frame rate to 60 fps during gameplay whether or not VSync is enabled. Performance characteristics of a timedemo do not reflect actual gameplay. So why do we do them? Because the questions we are trying to answer have only to do with the graphics subsystem. We want to know what graphics card is better at rendering Quake 4 frames. Any graphics card that does better at rendering Quake 4 frames will handle Quake 4 better than another card. While that doesn't mean the end user will necessarily see higher performance throughout the game, it does mean that the potential for seeing more performance is there. For instance, if the user upgrades CPUs while keeping the same graphics card, having higher potential GPU performance is going to be important.

What this means to the end user is that in-game performance will almost always be lower than timedemo performance. It also means that graphics cards that do slightly better than other graphics cards will not always show a tangible performance increase on an end user's system. As long as we keep these things in mind, we can make informed conclusions based on the data we collect.

Our benchmark consists of the first few minutes of the first level. This includes both inside and outdoor sections, with the initial few fire fights. We tested the game with Ultra Quality settings (uncompressed normal maps), and we enabled all the advanced graphics options except for VSync. Id does a pretty good job of keeping framerate very consistent, and so in-game framerates of 25 are acceptable. While we don't have the ability to make a direct mapping to what that means in the timedemo test, our experience indicates that a timedemo fps of about 35 translates into an enjoyable experience on our system. This will certainly vary on other systems, so take it with a grain of salt. The important thing to remember is that this is more of a test of relative performance of graphics cards when it comes to rendering Quake 4 frames -- it doesn't directly translate to Quake 4 experience.

Quake 4 Performance


At resolutions over 1280x1024, the GeForce 7950 GT falls short of the X1900 XT 256MB. Even with all these cards offering playable performance at 1920x1440, it's still a better idea to stick with the card that will give you higher performance for less money. In this case, that would be the Radeon X1900 XT 256MB. That is, if you can get ahold of it at a reasonable price.

Quake 4 Performance


While these numbers are a little more cluttered with lower end SLI solutions falling on top of high end single card setups, the lead ATI has held since their release of Catalyst 6.8 remains strong, especially with AA enabled. Not only does the X1900 XT 256MB outperform the 7950 GT, it outperforms the 7900 GTX. Not even SLI is enough to give an advantage to NVIDIA in these tests, and the 7950 GX2 only barely edges out the X1950 XTX in performance.



Splinter Cell: Chaos Theory Performance

We make use of the Lighthouse demo for Splinter Cell: Chaos Theory. We have been using this benchmark for quite some time and facilitate automation with the scripts published at Beyond 3D. This benchmark is fairly close to in game performance for our system, but midrange users may see a little lower real world performance when tested with a lower speed processor.

Our settings all used the highest quality level possible including the extra SM3.0 features. As the advanced shaders and antialiasing are mutually exclusive under SC:CT, we left AA disabled and focused on the former. We set anisotropic filtering to 8x for all cards.

For this 3rd person stealth game, ultra high frame rates are not necessary. We have a good playing experience at 25 fps or higher. There may be the framerate junkie out there who likes it a little higher, but our recommendation is based on consistency of experience and ability to play the game without a degraded experience.
Splinter Cell: Chaos Theory Performance


Those out there who like a smoother than necessary framerate will be happy with the performance of Splinter Cell: Chaos Theory under any of these cards. Once again we see the X1900 XT 256MB card outpacing the 7950 GT, with 7950 GT SLI performance splitting the difference between the 7950 GX2 and 7900 GTX SLI.



Nothing much changes with the other resolutions. Even at 1920x1440 these cards remain playable. The 7900 GS pulling in at less than 30fps isn't that big an issue with this stealth based title, and 7900 GS SLI leads the X1950 XTX in performance. The 7950 GT lags behind both X1900 XT flavors throughout testing.



Final Words

The single 7950 GT card from NVIDIA does not seem to offer a greater immediate return on your investment than its major competitors in the X1900 XT family. First, the X1900 XT 256MB is cheaper than the 7950 GT. Both flavors out perform the 7950 GT in multiple games. Where the 7950 GT does lead the X1900 XT 256MB in a couple cases, it never leads the 512MB version. Obviously, the usual exceptions apply and those shopping with only one particular game in mind may form different conclusions, but in our eyes, the X1900 XT series are still better deals than stock speed 7950 GT parts.

Of course, there are already a number of overclocked 7950 GT cards out and about. These don't seem to be as highly overclocked as some of the 7900 GS cards we've been seeing around, but we are still planning on doing a 7950 GT mini-roundup in the next couple weeks in order to take a closer look at the differences in the overclocked options out there. If this card falls in price or if a highly overclocked version is available for the stock price, the 7950 GT may turn into a tempting deal. At this point, it just falls short in performance for the money. Of course, sometimes other factors matter more than performance in determining value. For instance, the passively cooled XFX 7950 GT cards provide zero noise with excellent performance.

The GeForce 7950 GT SLI can be a good deal compared to a $600 7950 GX2, but with the possibility of finding the GX2 for $550 and less (especially with mail in rebates) those who want the high end performance of SLI and the flexibility of a single card that can run in most motherboards will prefer the GX2. GeForce 7950 GT SLI is technically faster than a GX2 card, as the clock speeds on the individual GT cards are higher than those of the GX2, but performance is close enough that the GX2 is probably the better choice for SLI. The GeForce 7950 GT SLI option is probably better as an upgrade later on when the added performance becomes necessary.

In most cases, even the 7900 GS SLI setup performs as well as the X1950 XTX. If you don't need maximum performance right now, a $200 investment could offer a good performance upgrade for older GPUs, and you can eventually add a second card to get current high end single GPU performance for less money in the long run. You could also look at spending $400 to go straight to 7900 GS SLI, which will typically give you performance slightly better than a single 7900 GTX or X1950 XTX card at a lower price -- though of course you still need an SLI compatible motherboard.

With DirectX 10 and Microsoft Windows Vista on the horizon, we're hesitant to recommend dropping a lot of money on a GPU upgrade right now if you don't need it. DirectX 9 hardware should still remain useful for years after the DirectX 10 launch, but unless you really have a lot of disposable income and are willing to upgrade GPUs again in six months, our advice would generally be to stick with current generation cards costing $300 or less. Such cards offer more than enough performance for the vast majority of users, and even if DirectX 10 won't be a factor in the short term, DX10 class cards will be able to handle DX9 class games all the better. We expect high end cards built for DX10 to well exceed the performance of DX9 hardware even under current games.

Log in

Don't have an account? Sign up now