XFX 7900GTX XXX Edition Video Card Review
7900 GTX:
8 vertex pipes
24 pixel pipes
16 ROPs
650 MHz core clock
1600 MHz memory data rate
quote : On the high end, the 7900 GTX generally performs around the X1900 XT and X1900 XTX. This isn't a blow out victory for either NVIDIA or ATI as far as performance goes.
In general, SLI edges out CrossFire in most cases
anandtech has good scores, the same as guru, ..., and the same conclusion for a 650mhz card, bye alle have fun
8 vertex pipes
24 pixel pipes
16 ROPs
650 MHz core clock
1600 MHz memory data rate
quote : On the high end, the 7900 GTX generally performs around the X1900 XT and X1900 XTX. This isn't a blow out victory for either NVIDIA or ATI as far as performance goes.
In general, SLI edges out CrossFire in most cases
anandtech has good scores, the same as guru, ..., and the same conclusion for a 650mhz card, bye alle have fun
First of all welcome to the forums! Thanks for the questions, and feedback. 
The issue here is that we didn't follow the review guide sent out by Nvidia/ATI like pretty much every other site does. The two sites that you mentioned DO NOT use the same settings as Legit Reviews.
When reviewing high end video cards we like to test at settings and on equipment those gamers buying a $500-$650 video card would likely use and that isn't usually with the default driver settings or on a 1024x768 monitor. Not that there is anything wrong with 1024x768 gaming, but a $500 video card is a bit of a waste in that scenario. Likewise someone with a 30" Apple Cinema Display probably isn't going to be gaming with a X1800XL, they will be buying the highest end video card they can get their hands on.
Texture quality is very important in today's games, more so than it was in the past. Running the driver defaults on with Nvidia cards results in texture quality lower than ATI's, especially with ATI cards running HQ AF. This has been an issue on Nvidia hardware since the introduction of the 6800's.
We tested with a Dell 2405FPW, a 24" widescreen monitor running at 1920x1200, which amplifies deficiencies in texture quality.
Why would you buy a video card this expensive and then not run it to have the best image quality??
One last thing to note is that not all review sites use the same tests. Obviously Guru3d used a much less demanding test for Serious Sam 2 (113fps @ 19x12 with HDR on). We use a much more intense test which could change results between the cards as it may be a strong suit for one card or another. Typically the tests chosen are the most demanding that we can find, as this will give you an idea of worst case scenario, which is when you truly need the more powerful card.
The bottom line is that we don't test the cards to get the highest benchmark numbers, we test them the way that they should be and likely, will be used.
Again, thanks for the questions. Please feel free to give any suggestions that you may have. If there are other tests or settings you'd like to see we may be able to include that in future reviews where there are less time constraints.
The issue here is that we didn't follow the review guide sent out by Nvidia/ATI like pretty much every other site does. The two sites that you mentioned DO NOT use the same settings as Legit Reviews.
When reviewing high end video cards we like to test at settings and on equipment those gamers buying a $500-$650 video card would likely use and that isn't usually with the default driver settings or on a 1024x768 monitor. Not that there is anything wrong with 1024x768 gaming, but a $500 video card is a bit of a waste in that scenario. Likewise someone with a 30" Apple Cinema Display probably isn't going to be gaming with a X1800XL, they will be buying the highest end video card they can get their hands on.
Texture quality is very important in today's games, more so than it was in the past. Running the driver defaults on with Nvidia cards results in texture quality lower than ATI's, especially with ATI cards running HQ AF. This has been an issue on Nvidia hardware since the introduction of the 6800's.
We tested with a Dell 2405FPW, a 24" widescreen monitor running at 1920x1200, which amplifies deficiencies in texture quality.
Why would you buy a video card this expensive and then not run it to have the best image quality??
One last thing to note is that not all review sites use the same tests. Obviously Guru3d used a much less demanding test for Serious Sam 2 (113fps @ 19x12 with HDR on). We use a much more intense test which could change results between the cards as it may be a strong suit for one card or another. Typically the tests chosen are the most demanding that we can find, as this will give you an idea of worst case scenario, which is when you truly need the more powerful card.
The bottom line is that we don't test the cards to get the highest benchmark numbers, we test them the way that they should be and likely, will be used.
Again, thanks for the questions. Please feel free to give any suggestions that you may have. If there are other tests or settings you'd like to see we may be able to include that in future reviews where there are less time constraints.
Thanks for clarification.
My unassuming wish. As to applying antialiasing, could it be a bit easy to compare the boards at something like 6aa+ adaptive (for ati) (and (4msaa+2ssaa for nvidia resp.) striving for more stressed conditions, and at the same time using well known timedemos in a way that we( readers) could compare the results from the other sites to yours.
Me personally thinks your current method is fair enough, but the one thing bothers me here that if one IHV decided to choose not to disable all optimizations under HQ mode and hence made this testing methodology less meaningfull ( which i hope they won't do )
My unassuming wish. As to applying antialiasing, could it be a bit easy to compare the boards at something like 6aa+ adaptive (for ati) (and (4msaa+2ssaa for nvidia resp.) striving for more stressed conditions, and at the same time using well known timedemos in a way that we( readers) could compare the results from the other sites to yours.
Me personally thinks your current method is fair enough, but the one thing bothers me here that if one IHV decided to choose not to disable all optimizations under HQ mode and hence made this testing methodology less meaningfull ( which i hope they won't do )
-
Imakeholesinu
- Legit Aficionado
- Posts: 94
- Joined: Tue Mar 07, 2006 6:07 pm
- Location: St. Louis/Jackson, Missouri
-
Apoptosis
- Site Admin
- Posts: 33941
- Joined: Sun Oct 05, 2003 8:45 pm
- Location: St. Louis, Missouri
- Contact:
I personally keep all reference cards and one product from each chipset, gpu, cpu, memory line and so on for future articles. While we give away product on the site we will only give away retail products. The pre-production samples we have come with no warranty and are usually buggy and early revisions (motherboards are really bad). Also keep in mind that Legit Reviews has over six writers, a graphics designer, a site admin and a PHP coder to keep hooked up.Imakeholesinu wrote:When's he giving this one away?
-
Apoptosis
- Site Admin
- Posts: 33941
- Joined: Sun Oct 05, 2003 8:45 pm
- Location: St. Louis, Missouri
- Contact:
The CATALYST 6.3 drivers did make a difference in the benchmark scores as you could see. Our numbers are spot on and as I like to say "straight up and on the money". This is one review site where the numbers don't have to be questioned. Legit Reviews does test the cards unlike some sites. Many review sites just don't take the time to benchmark with updated drivers on launch day. Within a week of the launch ATI released the Beta CATALYST 6.3 drivers and NVIDIA gave us 84.11's then 84.12's then 84.14, and then finally 84.17's. Wickedld9 benchmarked his heart out for you guys to give you the right numbers on launch day and had to fully re-bench the cards several times and update his numbers. I personally would like to thank Wickedld9 for busting his ass and giving the enthusiast community some of the only current 7900 GTX, 7600 GT, and ATI X1800GTO numbers on the net.DevilCry wrote:All the others used 6.2 drivers for ati.On this review they used 6.3.Is this answer fair enough for you?mister X wrote:aah thats why every other review has different scores, thanks for the explenation lollll
Oh and welcome to the forums DevilCry and great first post!
