ATI vs nVidia TRAOD benchmark

[simulation]

Tom's Hardware Guide have just published a review of the new ATI Radeon XT and they have used TRAOD as one of the benchmarks (one of the first DX9 Game).
The full report can be read here at Tom's Hardware Guide (http://www4.tomshardware.com/graphic/20031015/radeon9600xt-16.html), but basically the ATI Radion cards are out performing the equivalent nVidia FX cards by at least 2:1 and that is even against the new v51.26 drivers

[Karthik]

Not good! NVIDIA is in hot water!

[fastclick]

There already was a post about this look down this verry page.
In my opinion nVidia is way over ATI. The cards are cheeper and if Tom's Hardware Guide would test would the cards on older computers (no one here has a P4 3.2Ghz) like a 1.8Ghz or 2.6Ghz etc. there would be many many many differences.
It's all about your comp not about the guide.

[simulation]

Hi Fastclick,

I'm compiling a list of graphic card performance ratings. Do you fancy running the TRAOD benchmark?

There are instructions Here (http://www.simulations.btinternet.co.uk/) on how to run the test and graphs showing the performance. I agree with you that all the official tests are run on the latest harware, so it would be nice to see some real results.

So far I have tried the Ti440, FX5200 and Radeon 9800Pro on a Athlon 1700+

The only snag is v52 patch has this feature removed (a nVidia request).

[steelmartin]

Hi all you wanderers in the graphic abyss,

on request, here's a link to a test - Ati vs Nvida in Tombraiderland - with an Athlon XP 'Thoroughbred' 2600+ (2.083 GHz):
FX 5700 : 52.16 vs 9600 XT: Catalyst 3.8 in TR:AOD (http://www.tech-report.com/reviews/2003q4/geforcefx-5700ultra/index.x?pg=10)

There's an interesting comment about the benchmarks under the graphs.

Simulation, someday when I find the time, :) I'll try to do your benchmarks... do you know how the game responds when you enforce PS 2.0? My hardware (Nvidia) doesn't support that...guess I'll have to remove those tests?

/ M

[fastclick]

Well, Maybe as I fix comps I have access to many video cards most of them from people that got new ones and dont need the older ones. I have an FX5900 its a realy good card. I also might get a Redeon 9800 but not sure yet. I can test the game on almost any old card :\ most are realy old...

[dhama]

Benchmarks aren't worth the paper they're printed on. We don't all see what's displayed on the monitor in the same way. Each one of us has subtle differences in what we see, so perhaps we should benchmark ourselves instead.
At the end of the day, the differences between video cards are negligible, and most of us won't notice any differences anyway; the only people that would are those that run benchmark software.

BTW, ATI cards are best :p

[simulation]

differences between video cards are negligible

You are joking arn't you?????:D

Most of the players complaining about the controls being sluggish, awkward to use, the game slowing down in certain sections (graveyard, aquatic research area) can blame it on their old graphics cards.
I work for a Software company testing OpenGL applications and get to use a wide variety of cards. So when I found TRAOD was running a bit slow on my Ti4400 I decided to try the FX5200. This gave me the PS2.0 options, but wasn’t much better, so I then swapped it out for a ATI 98000 Pro………. And :p:pWOW:p:p suddenly TR took on a whole new life. I can turn on all the extra DX features and still run at high FPS.

See the graph below. Well the 9800pro was on a Athlon 1700+ while the FX5200 was on a P4 3.06GHz. So I guess that shows what a difference a card makes.

http://www.simulations.btinternet.co.uk/data/1024Compare.gif

I know it isn't fair comparing to nVidia comparing the FX5200 with the 9800Pro, but I don't have that data yet. I should think nVidia 5600 up or ATI 9600 up would be OK

Anyone want to try the test?

[dhama]

Originally posted by simulation
You are joking arn't you?????:D


Not at all..... I can remember my 4mb Oak technolgies video card running the Original Doom, and boy was I taken a'back. Lets face it, video cards are as good the the best one you've seen so far.....

[Karthik]

Originally posted by dhama
Not at all..... I can remember my 4mb Oak technolgies video card running the Original Doom, and boy was I taken a'back. Lets face it, video cards are as good the the best one you've seen so far.....
Eh, DOOM could run on 1MB!:p

[dhama]

Originally posted by Karthik
Eh, DOOM could run on 1MB!:p

yeah..... well I was running four versions of it on the same PC. ;)

seriously though, as you can see, I had a cutting edge card at the time..... a bit like driving a porshe on a sunday.

[simulation]

Hi Steelmartin,

Simulation, someday when I find the time, I'll try to do your benchmarks... do you know how the game responds when you enforce PS 2.0? My hardware (Nvidia) doesn't support that...guess I'll have to remove those tests?

I have tried out the GeForce2 GTS which doesn't support PS2.0 and it just terminates immediatley and continues with the rest of the benchmark. So there is no need to edit the batch file.

Latest benchmark reseults can be foud HERE (http://www.simulations.btinternet.co.uk/) and I have added a number of images showing the difference in quality between older and latest cards.

[jack_Raider]

simulation,

you tested AOD on a Nvidia Quadro FX500? Aren't the Nvidia Quadro cards geared towards 3D modeling/animation?

It terms of performance and image quality, how well did AOD run on the Quadro? Was there any slow-down, missing textures, choppy/sluggish control?

[simulation]

Hi jack_Raider,

You are quite right that the Quadro is aimed at the OpenGL CADCAM market. I work for a CADCAM software company on the QA side, so I get to test a fair number of cards. I happened to have a Quadro FX500 in my draw, so I thought I would give it a go.

I found the Quadro FX500 (bottom end of the new Quadro FX range) slightly better than the GeForce FX5200, but not by much. Considering the cost difference £60 compared to £275 I think I can safely say gamers won't be running down to the shops.

You can see my comparisons HERE (http://www.simulations.btinternet.co.uk/)

What I really need is a cross section of common cards run on different CPU's to build up a picture of how TRAOD runs on different harware. This data can help gamers decide if a new card will solve their problems before splashing out hard earned cash.

[Zeph]

I played TRAOD with my Geforce 2 MX 400 and in the house of a friend with a Radeon 9800 Pro 128 and the game was equal in my PC, was only faster when loading the game because it has a Pentiun IV 2.1GHz...... (my english is bad XD)

[simulation]

Hi Zepth,

I did try a GeForce2 GTS and the frame rate was OK, but it doesn't have any support for the Pixel Shaders so the extra effects are not used like Shadows, Depth of Field, reflective water etc. and the Radeon can increase the screen resolution without much loss in performance. However, the Radeon 9800 Pro is an expensive upgrade and some would wonder if the extra effects are worth the upgrade?