Which video card??

[hl2maps.net]

An email i got from pcstats.....

RADEON 9800 PRO vs. NVIDIA FX5900

Hello peter,
Wow! PCstats has some really exciting reviews for you to read in this weeks Edition! In the left corner we have the ATI Radeon 9800 Pro, and in the right - nVidia's GeForceFX 5900 Videocard from newcomer Albatron. Who will be the winner, and who will be the vanquished? When it comes to a battle of GPU vs. GPU on the field of FPS, your court side seats are the best in the house!

What good is a fast videocard if your current motherboard goes back to the day of the BX chipset? Not much unfortunately, but there is a good upgrade path in the form of the i865PE chipset. We had some reviews of Springdale-based motherboards in the last few PCstats.com Newsletters, so today we are examining an i875P motherboard from EPOX called the EP-4PCA3+. PCstats.com recently tested the Asus P4C800 DLX 800MHz FSB Pentium 4 motherboard which is also built around the powerful i875P chipset, so be sure to read that review if you are deciding between i865PE and i875P chipsets.

I'm looking forward to what could be the 'Pentium 5' processor, and in this week's High Tech Low Down Chris fills in the the details as they have started to emerge. Colin talks memory speed in the Tech Tips which are always helpful, and remember to visit www.PCstats.com for up to the minute news throughout the day!

Thanks for reading the Weekly PCstats.com Newsletter,
Max Page
The Editor


The article is at:

http://www.pcstats.com/articleview.cfm?articleid=1436&page=1

 

[Guest]

Go for PowerColor or Saphirie Radeon 9800 pro with 256MB.

FX got many problems such as DX9.0,

oh ...NVIDIA goin to release their next driver this october FX 50.something

 

[Loke]

I would go for an ATI Radeon 9800.

 

[Sgt.Igneri]

the 5900s raw performance is better then the 9800s, but when AA and AF are turned on, the 9800 is ahed, and it has better image quality.

 

[DeathlessOne]

Chances are the problems with NVIDIA's 5900 and Hl2 will be solved by its release. Both cards are fine, but I'm staying with the 5900 Ultra.

...I am the NVIDIA whore...

 

[Ridic]

if this turns out to be another flame "ati vs nvidia" thread im going on a killin spree.



get a 9800 pro unless for some reason you want a 5900 ultra, the 9800 pro is better




PERIOD

 

[PsychoEva01]

FX 5900 is better than the Radeon 9800 in every way, but id get the 9800 because its the best deal (Price to Power)

 

[simmo]

AA Decreses performance right ?

 

[Killj00]

your wrong physcoeva...the FX and TI cards are having problems with DX9, those 2 cards are based on a GPU engine and the RADEON cards are based on VPU engines therefore it will be easier to get by the problem with DX 9... when FSAA is enabled on the nvidia card,image and graphics are all screwed up for hl2...i dont have a 9800 but i do have 9700 and i aint experiencing the best cause my LCD only supports 1024 by 780...i'm thinkin of gettin a new monitor...

 

[EVIL]

and besides that, even Gabe said that it was a better choice to buy a radeon 9800 pro for halflife2 (read one of the stickythread containing e-mails to gabe)

 

[Killj00]

hl2maps.net, if u wanna get the 9800,there are 2 versions...1 which only has 128 mb ddr which saves u money and the other 256 mb ddr which is of course more expensive...but even the 128 one is good enough, it wont last as long as in that the games in the future will need more memory then that but it's worth it and will last for a while cause i have yet to see a game that requires 128 mb...i have seen 64 but thats it;...

 

[dawdler]

Originally posted by Killj00
hl2maps.net, if u wanna get the 9800,there are 2 versions...1 which only has 128 mb ddr which saves u money and the other 256 mb ddr which is of course more expensive...but even the 128 one is good enough, it wont last as long as in that the games in the future will need more memory then that but it's worth it and will last for a while cause i have yet to see a game that requires 128 mb...i have seen 64 but thats it;...

256mb is definetly not worth it. By the time we see games using it, the card will be obsolete for most people (not budget buyers). Games that will require 256mb will probably require badass cards anyway. Just look at the R420, if its all its hyped up to be (and it does seem so) it will be at least TWICE as fast as the 9800. It would mean a significant jump, and all the other card manufacturers would have to follow. So then you start coding for 256mb, and then you assume the engine will be out in 2-3 years, when the R420 speeds are in every mans house, meaning the 9800 would be very slow. If the R420 isnt, its the NV40, it doesnt matter. Whichever is fastest WILL set the pace.

Short point: Its too soon to invest in 256mb.

Back on topic: Lets face it, the 9800 Pro has FAR superior DirectX9 speeds. Its been proven over and over and over again. The 5900 Ultra on the other hand has superior (but not by far) DirectX7-8 speeds. It is very simple to make a choice on NEW hardware that you want to buy. You do not go around buying OLD hardware for a NEW computer, not usually anyway (and especially not in the price ranges we are talking about).

 

[Pr()ZaC]

The 5900 Ultra on the other hand has superior (but not by far) DirectX7-8 speeds.

^^^with AA/AF off.

 

[SidewinderX]

can someone link me to these nVidia Dx9 problems? I have heard the Radeon AIWs had dx9 probelms, but I haven't heard anyhting wrong with nVidia....

 

[dawdler]

Originally posted by SidewinderX143
can someone link me to these nVidia Dx9 problems? I have heard the Radeon AIWs had dx9 probelms, but I haven't heard anyhting wrong with nVidia....

What, have you missed the entire thing?

We did a simple test here, hidden in the forums somewhere, a demo that simulate real time lighting using PS2.0. The 5900 Ultra was massing a whopass 15 fps without AA/AF. My 9700 Pro can do it at 30 fps with 6xAA/16xAF... Also, for instance [T]ardOCP's controversial BFG (or whatever it was) 5900 Ultra test (along with all the other tests on every page), shows it lagging behind 50-100% in DX9 test in shadermark. Except when using integers and displacement bump mapping or something, where it can win, by alot even. (5/30 tests or something)
Then of course we have HL2, where the devs pick the 9800 due to the 5900 terrible DX9 performance, they've said it themselves

 

[Sgt.Igneri]

Originally posted by dawdler
[T]ardOCP's

LOL! Was that serious? Cause Hard OCP i always found to be good.

 

[SidewinderX]

controversial BFG (or whatever it was) 5900 Ultra test (along with all the other tests on every page), shows it lagging behind 50-100% in DX9 test in shadermark.

I don't know what you'r tlaking baout....

From [H]ard OCP

In-Game Still Screenshot No AF = 5900 Ultra SLIGHTLY Sharper Textures with no mipmap transitions that are distracting.

In-Game Still Screenshot 8XAF = 9800 Pro SLIGHTLY Sharper Textures with no mipmap transitions that are distracting.

In-Game Movement No AF = Can’t tell any difference.

In-Game Movement 8XAF = Can’t tell any difference.

[quoye]We did a simple test here, hidden in the forums somewhere, a demo that simulate real time lighting using PS2.0. [/quote]
Key word there is PS2.0. No game even in development pretends to use or think baout using ps2.0. They use ps1.4, also a dx9 pixel shader. This is the shader that the FX series was build to use, and it is what games use. the only program that uses ps2.0 is 3DMark03, which is another reason why 3DMark03 is crap (see my thread) A ps2.0 tester only refelcts ps2.0 performance, which is not in any game, therefore it doesn't reflect on gaming performancone in the least.

 

[dawdler]

Originally posted by Sgt.Igneri
LOL! Was that serious? Cause Hard OCP i always found to be good.

Has always been Nvidia biased, and that Brent dude that did the review seem to have the same knowledge of hardware as a bacteria on mars does. Not to mention, reading up on all his replies in the forum, he appears to be a fine grade A jerk

Just an example is the test I mentioned, using DX9 PS2.0 for simulating realtime lighting. Very simple test. His card scored 5-10 fps! (I was actually impressed by the dude with the 5900 Ultra that got 15 fps, I expected it to be much lower). Now, did he test the 9800 Pro with it? No. Did he comment on the low performance? No. Did he say it might be a problem with PS2.0? No. Did he say anything at all about it? No. Just 5-6 lines and 3 small images at the very end of the review. He didnt even WRITE the results, just displayed them on the images (which you had to click to see results).
But...
The 5900 Ultra WINS IN THE ALLMIGHTY UT2K3!!!! This was displayed on MULTIPLE pages. And the issue with UT2k3 and Nvidia was LONG known, about it cutting corners, and can never acheive trilinear. Still, not a word on it. Not a single word.

Edit: Just saw the above reply

Key word there is PS2.0. No game even in development pretends to use or think baout using ps2.0. They use ps1.4, also a dx9 pixel shader. This is the shader that the FX series was build to use, and it is what games use

PS1.4 is DX8. And its old. PS2.0 is alot better. HL2 will without a doubt make use on it on DX9 cards. Besides, the FX can also do the PS2.0+... You think they meant for it to be PS1.4? That's the G4 dude. FX is meant to be Cg.

 

[GordN FreeLoadR]

can someone link me to these nVidia Dx9 problems?

Take your pick...

Article on nVidia "optimizations"
http://www.extremetech.com/article2/0,3973,1088795,00.asp
http://www.extremetech.com/article2/0,3973,1105259,00.asp

Article on nVidia "optimizations"
http://www.digit-life.com/articles2/gffx/gffx-13.html

Further NVIDIA optimizations for 3DMark03?
http://www.tech-report.com/etc/2003q2/3dmurk03/index.x?pg=1

Article using the RivaTuner guy's anti-cheat scripts
http://www.beyond3d.com/misc/random/antidetector/

Video Card Review Commenting on nVidia's "optimizations"
http://www.digit-life.com/articles2/gffx/gffx-16.html

Thread at Beyond3D About nVidia Not Doing Real Trilinear Filtering
http://www.beyond3d.com/forum/viewtopic.php?t=6719

The Truth About nVidia's Driver Performance
http://www.amdmb.com/article-display.php?ArticleID=247&PageID=1

Nvidia confirms PS 2.0 driver problems
http://www.theinquirer.net/?article=10602

NVIDIA-based Graphics Cards Will Have a Bug in Half-Life 2?
http://www.xbitlabs.com/news/video/display/20030718155730.html

nVidia's Risky 3D Optimization Gamble
http://www.extremetech.com/article2/0,3973,1201076,00.asp

As the GPU world spins...
http://www.tech-report.com/onearticle.x/5418

DirectX 9 Benchmarks ( Test it yourself )
http://esprit.campus.luth.se/~humus/
http://www.daionet.gr.jp/~masa/rthdribl/


G.F.

 

[dawdler]

http://esprit.campus.luth.se/~humus/

Anyone have a 5900 Ultra? Do the test with PS/VS 2.0 and 1.4 and etc, I can do them on my 9700 Pro (or if anyone got a 9800 Pro).
It might be fun to see the results
I dont care who wins really (well, I'm betting on my 9700Pro, but the 5900U might surprise us ).

Note: I havent downloaded any, but one problem might be they are both as fast as they seem simple, meaning they will hit the roof

Edit2:
Oh and I forgot, if all current games use PS 1.4, then why does the PS heavy parts of Splinter Cell HEAVILY outperform the 5900U? SC is faster on the 9800 in general, but when you get to the PS areas, the 9800Pro zooms past the 5900U... PS 1.4? You said it

 

[SidewinderX]

*NOTE*
I am hereby making an apology to everyone. I have mistakenly reversed the pixel shaders. 3DMark03 focus on ps1.4, whioch is not used by any games. most current games use ps1.3 and mst new one uses ps2.0, which 3Dmark03 doesn't test. I am very sorry.
*NOTE*

Anyone, about that whole "cheating" things...

http://www.futuremark.com/news/?newsarticle=200306/2003060305#200306/2003060305

Futuremark Statement
For the first time in 6 months, as a result of Futuremark's White Paper on May 23rd, 2003, Futuremark and NVIDIA have had detailed discussions regarding NVIDIA GPUs and Futuremark's 3DMark03 benchmark.

Futuremark now has a deeper understanding of the situation and NVIDIA's optimization strategy. In the light of this, Futuremark now states that NVIDIA's driver design is an application specific optimization and not a cheat .

The world of 3D Graphics has changed dramatically with the latest generation of highly programmable GPUs. Much like the world of CPUs, each GPU has a different architecture and a unique optimal code path. For example, Futuremark's PCMark2002 has different CPU test compilations for AMD's AthlonXP and Intel's Pentium4 CPUs.

3DMark03 is designed as an un-optimized DirectX test and it provides performance comparisons accordingly. It does not contain manufacturer specific optimized code paths. Because all modifications that change the workload in 3DMark03 are forbidden, we were obliged to update the product to eliminate the effect of optimizations identified in different drivers so that 3DMark03 continued to produce comparable results.

However, recent developments in the graphics industry and game development suggest that a different approach for game performance benchmarking might be needed, where manufacturer-specific code path optimization is directly in the code source. Futuremark will consider whether this approach is needed in its future benchmarks.

 

[Ender01]

Chances are the problems with NVIDIA's 5900 and Hl2 will be solved by its release. Both cards are fine, but I'm staying with the 5900 Ultra.

Kevin, Kevin, Kevin....
there is a work-around, but they have to emulate a type of texture compression that Ati has built in, this emulation slows down the proformance! Ati is cheaper, and faster! They need no slogan.

 

[Ridic]

Originally posted by PsychoEva01
FX 5900 is better than the Radeon 9800 in every way, but id get the 9800 because its the best deal (Price to Power)

omfg you douche bag.

im nto even going to point out all the things you said wrong, mainly because i dont want this thread to be 75 pages long.

but i will tell you that you need to rethink your life. Go somewhere for the weekend to get away from everything. When you change, come back. Cause right now your a litte 11 year old that spues out any iformation he wants, true or not true...people like you are the ones who start all the "blank vs blank" threads that usally end up being closed on most forums. So i suggest you just....leave for a little while..

 

[dawdler]

*NOTE*
I am hereby making an apology to everyone. I have mistakenly reversed the pixel shaders. 3DMark03 focus on ps1.4, whioch is not used by any games. most current games use ps1.3 and mst new one uses ps2.0, which 3Dmark03 doesn't test. I am very sorry.
*NOTE*

Actually, there is a both a PS 2.0 and VS 2.0 in Mother Nature. Not much, but its there. Not enough to be a true 2.0 bench, that I can agree on.