Dear forum members,
We are delighted to inform you that our forums are back online. Every single topic and post are now on their places and everything works just as before, only better. Welcome back!


Discussion on Article:
The Fastest Graphics Cards of Summer 2004

Started by: lol | Date 08/01/04 08:24:29 PM
Comments: 80 | Last Comment:  12/09/16 10:47:12 AM

Expand all threads | Collapse all threads


Note to the author, regarding your IL-2:FB tests. This sim plays faster in OpenGL with better graphical quality and in that regime the nVidia cards equal and better their ATI counterparts easily. It is obvious from your numbers that you tested in Direct3D. Just thought you should know.
0 0 [Posted by:  | Date: 08/01/04 08:46:55 PM]
- collapse thread

In fact, we used OpenGL render-path for IL2.
0 0 [Posted by:  | Date: 08/03/04 02:43:30 PM]
Just to reiterate. I test FB all the time because it is my main pass-time. On my meager rig (P4 2.4C, 1GB PC3200 RAM and Radeon 9700, non-pro) I average 26fps on the highest possible settings (perfect landscape) in OpenGL, black-death track, at 1024x768x32 + 4xFSAA and no Aniso. Explain to me how I can get almost double the frame rates of the Radeon 9800XT at 1024x768 with my aging rig? Explain to me how every other FB player is getting comparable results to mine? I'm sorry but your numbers just don't jive.

If you are unaware, "The Black Death" track is the most widely used benchmarking utility in the FB world because of it's incredible load on both the CPU and the graphics card.
0 0 [Posted by:  | Date: 08/03/04 02:55:36 PM]
We did not use "The Black Death" track, which is why you should not compare your results to us, especially keeping in mind that we have totally different system compared to yours.
0 0 [Posted by:  | Date: 08/06/04 12:53:46 PM]
Look, I'm not a fan of either brand of video card and don't care who's on top. But this is an obviouse case of testing error. I don't care what track you're using there's something fishy with your results. As I said, they are lower than the numbers I'm getting on my low end rig on one of the most demanding demo tracks in FB.

I can't understand your arrogance? Everyone else has posted completely different numbers whether they be other online reviewers or other gamers using FB and yet you refuse to even entertain the idea that something is wrong with how you tested. Incredible! I'll leave you to your ATI-worship - at least I now know where you stand. B'Bye now...
0 0 [Posted by:  | Date: 08/08/04 11:17:28 PM]

First off, well done on doing such a large and exhaustive review that no doubt took a long time.

However, there is a flaw. Why must Xbitlabs ALWAYS neglect to mention that Nvidia's driver optimizations are turned OFF?

You cannot turn off ATI's optimizations, and so ATI gets the edge once you turn on AA+AF.

If you were to re-run the benches with Nvidia's optimizations ON, then Nvidia would have looked alot better with AA+AF..

Atleast edit your review and put the important information of whether Nvidia's optimizations are on or off in the test bed section..
0 0 [Posted by:  | Date: 08/02/04 08:00:36 PM]
- collapse thread

We enabled trilinear and anisotropic filtering optimizations for NVIDIA GeForce cards.
0 0 [Posted by:  | Date: 08/03/04 02:16:47 PM]
Which tests?
0 0 [Posted by:  | Date: 08/04/04 12:15:49 PM]

I'm impressed by the scope of the article. Very well done. We need articles like this. :)
Unfortunately though, this article was far too biased to be taken seriously.

Whenever NVidia wins, we get a comment like "NVIDIA’s solutions are evidently better in this game, but it’s hard to account for this fact." and "At least we can’t lay blame on the low geometry processing speed anymore – ATI’s new graphics processor is NVIDIA’s better in this respect as you know."
What we *know* is that NVidia won that benchmark. Trying to pass that off as some anomaly, while stating (without any proof) that ATI's cards are better in this respect, just doesn't sound very professional. If they are better, show us. If you can't show us, don't pretend that it's a fact.

Other comments are "With FSAA and AF enabled, NVIDIA’s graphics cards look unsure, as we’ve seen a number of times, since their algorithms of working with the memory are less efficient." about a benchmark where the 6800 U beats the X800XT with 10+ fps. If that is unsure, then what is ATI's cards? Paralyzed?

But when ATI cards are ahead, there's no need to look for an explanation.
In another, we're told that the 3-5 fps difference (in NVidia's favor) is a "small difference", but a few pages earlier, ATI is said to perform better because it has about the same number of fps over the NV cards.

"The RADEON X800 XT equals the GeForce 6800 Ultra in this game" about a benchmark where the 6800 Ultra wins with 8/1/4 fps (for each resolution)
We're talking about something like a 10% lead to NVidia here.

"The top-end members of the new GPU series from ATI and NVIDIA again give out the same number of frames per second" in Flight Simulator 2004, where the 6800 U wins by 3-4 fps in all resolutions and in pure as well as eye candy, but on the next page (X2), "The RADEON X800 XT seems to be winning in this test", despite only having a lead of 0.6, 1.5 and 0.6 fps in the different resolutions.
In Eye Candy, the 6800 U pulls ahead by 7, 3 and 1 fps, but we're told that "The situation hardly changes in the “eye candy” mode"

I'm not questioning the benchmarks, and I'm not saying ATI's cards sucks, I'm asking for a fair description of the benchmarks. I'm still trying to decide which card to buy, and reading biased articles like this doesn't help. I want the best card, not the one from the reviewers favorite company.
0 0 [Posted by:  | Date: 08/03/04 08:46:56 AM]

Anton you need to address the fact that Nvidia's driver optimizations were turned OFF! You;re the only review site I've ever seen in which ATI beat Nvidia in CoD, which is an OpenGL game!

The reason for that is because Nvidia's AF optimizations were disabled, which isn't fair.
0 0 [Posted by:  | Date: 08/03/04 07:46:49 PM]

Disregard that last comment by me please... I just read that you said you did enable driver optimizations..

However, I find that weird because the NV40 always beats ATI in CoD.. If you look at some of your past reviews, you can see this clearly...
0 0 [Posted by:  | Date: 08/03/04 07:48:58 PM]

One mistake in your Microsoft flight simulator 04' bench's, is that the last eye candy mode benchmark, shows the same benchs as the non eye-candy mode, if someone could fix that with the eye candy mode benchs, it would be helpful.
0 0 [Posted by:  | Date: 08/04/04 07:35:27 AM]

>>If silence, compactness, power-saving are notions that matter much to you, or if you’re a real hardware enthusiast, consider the RADEON X800 Pro. Modders should be interested in this product since it can be converted into a ******RADEON X800 XT Extreme Edition****** with a bit of skill and luck. In this case you’ll get a nice performance boost. ATI’s R420 chip yield is very high and some of fully operational dies go to produce 12-pipeline cards. This is the ground for the conversion, although you should approach the problem soberly and not risk it if you don’t feel confident as to the outcome. !!
0 0 [Posted by:  | Date: 08/04/04 08:38:49 AM]
- collapse thread

Errr, my whole message didn't go through? Anyway, I hope you mean platnium edition and not extreme edition.

0 0 [Posted by:  | Date: 08/04/04 08:39:59 AM]

I love xbitlabs, but there is no way you guys are getting 50-70 fps in farcry with a 9800 pro/128 and all settings at the highest on 1024x768. Similar remarks in regards to all your other results
0 0 [Posted by:  | Date: 08/04/04 04:21:02 PM]

your high end graphics results r pretty useless since the rest of the system is causing some sort of bottleneck
0 0 [Posted by:  | Date: 08/10/04 03:24:40 AM]

please languange in german!!!!

0 0 [Posted by:  | Date: 08/11/04 02:40:01 AM]


Back to the Article

Add your Comment