bkswaney's GeForce FX 5800 Ultra report

[marcocom]

hovz wants to use that 256bit and get us up in the nose-bleed resolutions with full driver effects on.

want to talk about our hardware in terms of whether or not your 256bit solution is superior, then we strip ALL THE DRIVER LEVEL FEATURES like FSAA, AF and we let them rip. thats how benchmarking has always been done. fillrate. lowest mipmapping, no Vsync. old school. the way you would play it in a ladder match, to win.

ud be lucky to win half of them with the 9800 and this is a brand new architecture with new drivers and alot of ground and flexibility to explore.

or how about this, we just run the benchies and either of us can do whatever we want to the drivers to get the most speed, how about that... beat the FX in every bench?

[GlowStick]

Originally posted by Sazar
reading the same link I can't see where they say they TRY to help :)

they claim they are actively promoting their own kits in order for support of their cards... 2d/3d info available...

I personally don't use linux... hence I have not experienced the problems...

perhaps over the summer break I will hook up my k6-2 rig thats gathering dust and see if I can learn some 'nix :)

unfortunatley I don't have an agp slot on the k6-2 mobo :(

Linux will run on a 486... and less : )

[StealthHawk]

Originally posted by marcocom
hovz wants to use that 256bit and get us up in the nose-bleed resolutions with full driver effects on.

want to talk about our hardware in terms of whether or not your 256bit solution is superior, then we strip ALL THE DRIVER LEVEL FEATURES like FSAA, AF and we let them rip. thats how benchmarking has always been done. fillrate. lowest mipmapping, no Vsync. old school. the way you would play it in a ladder match, to win.

of course we aren't still stuck in those days where it is hard to maintain 60fps with all effects on.

and in most games where framerate is a problem, it is due to the games being CPU dependent, in which case stressing your 3d card out won't degrade performance much.

the days of playing without FSAA or AF are coming to an end. each new generation of cards that comes out makes it increasingly hard to justify not using FSAA, for example. the performance hit keeps getting lower and lower.

[Solomon]

Originally posted by StealthHawk
of course we aren't still stuck in those days where it is hard to maintain 60fps with all effects on.

and in most games where framerate is a problem, it is due to the games being CPU dependent, in which case stressing your 3d card out won't degrade performance much.

the days of playing without FSAA or AF are coming to an end. each new generation of cards that comes out makes it increasingly hard to justify not using FSAA, for example. the performance hit keeps getting lower and lower.

Especially when you are forking over $400.00 for a card, why wouldn't you want the ability to run the games smoothly with AA/AF? :D

Regards,
D. Solomon Jr.
*********.com

[hovz]

Originally posted by marcocom
hovz wants to use that 256bit and get us up in the nose-bleed resolutions with full driver effects on.

want to talk about our hardware in terms of whether or not your 256bit solution is superior, then we strip ALL THE DRIVER LEVEL FEATURES like FSAA, AF and we let them rip. thats how benchmarking has always been done. fillrate. lowest mipmapping, no Vsync. old school. the way you would play it in a ladder match, to win.

ud be lucky to win half of them with the 9800 and this is a brand new architecture with new drivers and alot of ground and flexibility to explore.

or how about this, we just run the benchies and either of us can do whatever we want to the drivers to get the most speed, how about that... beat the FX in every bench?

marcocom u are now the most retarded poster on this forum. u spend 10 pages trying to argue how speed isnt important its image quality, now u tell me that ur geforce is faster when u make the image as low quality as possible? isnt this suppose to be the age of cinematic rendering? but your suggesting the card that sacrafices image quality but performs the fastest is now the best? if your payying 400 dollars didnt u say u want ur money to be invested in better image quality? not just winning bnechmarks? also i sugeest u read the hardcop review where they use nivida application v ati quality and see the 9800 pro smoke the nvidia card in every single benchmark except serious sam. but oh ya ur card is faster in agressive mode. damn nvidia sure spent ur money where u wanted it, a card that excells in low image quality settings. cinematic gaming at its best

[jAkUp]

to hovz:

This card kicks @ss.

The effects even seem better than my 9700 was

this is a first hand account from bkswaney! why do you compare screenshots that use old drivers! dont compare the cards unless you have seen both of them in motion personally.... it makes no sense. i have a friend with a 9700, so therefore i have seen both in person, so i make a conclusion on which i think looks better. and to tell you the truth, they look identical to me. i know its an alpha and all.. but my card definetely had the doom3 alpha running and looking better. everything else looks exactly the same to me.

[GlowStick]

Originally posted by Solomon
Especially when you are forking over $400.00 for a card, why wouldn't you want the ability to run the games smoothly with AA/AF? :D

Regards,
D. Solomon Jr.
*********.com

Well, i thought that when i bought my Raden 9800Pro with its improved FSAA and AA core tweaks. But i have a probem, i like to play games at 1280x1024. Not pisses me off more then when i have to 'change resolutions'.

The naked truth is, you cannot play Unreal II with AA and AF maxed, nor can you play Unreal II with AA 4x and AF 8x. The final mode i can play Unreal II at a decent fps is AA 2x AF 4x, but i have to turn it off on the out door maps.

Im sure this post will recive the standad "you installed your card wrong" but i really do laugh at people when they post that.

[bkswaney]

Originally posted by GlowStick
Well, i thought that when i bought my Raden 9800Pro with its improved FSAA and AA core tweaks. But i have a probem, i like to play games at 1280x1024. Not pisses me off more then when i have to 'change resolutions'.

The naked truth is, you cannot play Unreal II with AA and AF maxed, nor can you play Unreal II with AA 4x and AF 8x. The final mode i can play Unreal II at a decent fps is AA 2x AF 4x, but i have to turn it off on the out door maps.

Im sure this post will recive the standad "you installed your card wrong" but i really do laugh at people when they post that.

God no. I could have told u that.
UT brings the Pro's to a crawl using AA/AF.
I used 2X AA and 2XAF last night with my new card and everything turned wide open on the game.
It run smooth as silk. :D
No stuttering and all like my pro.

The more I use the Ultra the more I like it. :)

[GlowStick]

Originally posted by bkswaney
God no. I could have told u that.
UT brings the Pro's to a crawl using AA/AF.
I used 2X AA and 2XAF last night with my new card and everything turned wide open on the game.
It run smooth as silk. :D
No stuttering and all like my pro.

The more I use the Ultra the more I like it. :)


You got studdering on your 9700pro also? I get that a bit in UII and alot in Rallisport challange, i mean ALOT.

[nagus]

.... and the winner is ...


Radeon 9700PRO >> 4674 points in 3Dmark2003 @ 300/300 :D :D :D

poor NV35 ;)

[RobHague]

But it wont be clocked at 300/300 will it? ;) At least 500/500 but since its not DDRII - Im thinking maybe 600/400 to be worth while.

[nagus]

Originally posted by RobHague
But it wont be clocked at 300/300 will it? ;) At least 500/500 but since its not DDRII - Im thinking maybe 600/400 to be worth while.


shure, but i think the NV35 will have a hard time if ati (or one of its partners like FIC ) launch a Radeon 9800PRO with 256MB ram and a clockspeed of ~ 400/460Mhz

[RobHague]

Considering how much a 256MB card with DDRII would cost to produce i doubt NVIDIA will be worried. ;) But we are making comparisons of a card that isnt even out yet so lets just wait till NV35 hits the shores. :)

[5150 Joker]

Originally posted by marcocom
nah bro. the FX ultra on DDR2 smokes up a 9700 in everything except when a few driver-level effects are applied.

i agree that no one should upgrade from the 9700pro to anything but the nv35...but that old 9700pro aint got nothing on my 5800ultra. thats a myth and every new driver version proves it more and more.

but theyre certainly close as hell and the age of the 9700pro means that it was unquestionably the 'card to own' last year and this year in terms of value and just about everything. frankly, i wish i had purchased one when they had released. (except no dual-cpu support makes it impossible) i would be definitely waiting for nv35

however, it also works when a user like bkswaney just sells the card, gets roughly half the money to upgrade to the latest. (for me, it seems cheap. i mean...i spend alot more on snowboard equipment and that stuff only gets used 5 times a year!)


:bs: You got it backwards, the NV30 gets beaten by a 9700 pro in most gaming benchmarks.

[GlowStick]

Originally posted by nagus
.... and the winner is ...


Radeon 9700PRO >> 4674 points in 3Dmark2003 @ 300/300 :D :D :D

poor NV35 ;)

Hahahaha what a LIAR!!!!!!!!

You are the biggist ATi Fanboy i have ever seen.

Actually a Radeon 9800PRO at 300/299 (300 clock is not possible) gets 1559 3dmark2003's.

Lets see Poor Radeon ATi, getting stomped pritty good.

Now, if i were you i would want a link or something, good thing i am a 3dmark03 PRO user

http://service.futuremark.com/compare?2k3=655082

[5150 Joker]

Originally posted by GlowStick
Well, i thought that when i bought my Raden 9800Pro with its improved FSAA and AA core tweaks. But i have a probem, i like to play games at 1280x1024. Not pisses me off more then when i have to 'change resolutions'.

The naked truth is, you cannot play Unreal II with AA and AF maxed, nor can you play Unreal II with AA 4x and AF 8x. The final mode i can play Unreal II at a decent fps is AA 2x AF 4x, but i have to turn it off on the out door maps.

Im sure this post will recive the standad "you installed your card wrong" but i really do laugh at people when they post that.

Unreal II is a cpu hog though. It runs like a dog on most systems no matter what card you have.

[5150 Joker]

Originally posted by GlowStick
Hahahaha what a LIAR!!!!!!!!

You are the biggist ATi Fanboy i have ever seen.

Actually a Radeon 9800PRO at 300/299 (300 clock is not possible) gets 1559 3dmark2003's.

Lets see Poor Radeon ATi, getting stomped pritty good.

Now, if i were you i would want a link or something, good thing i am a 3dmark03 PRO user

http://service.futuremark.com/compare?2k3=655082

Wow then you must be a complete moron that cannot build a computer properly. I can confirm his score with my own: http://service.futuremark.com/compare?2k3=655196

R9700 using 3.2 cats modified to 9800. Core/mem set at 301/301 (3DMark reads them at 302/302) using R3D Tweak. No wonder you get stuttering and other issues, it's obvious your rig has problems due to a severe PEBKAC issue. :rolleyes:

For those too lazy to look at the compare links between mine and glowsticks:

General Information

Operating System Microsoft Windows XP Microsoft Windows XP

DirectX Version 9 9

Mobo Manufacturer ABIT ABIT

Mobo Model i845G-W83627HF i845G-W83627HF

AGP Rates (Current/Available) 4x / 1x, 2x, 4x 4x / 1x, 2x, 4x

CPU Intel Pentium 4 2408 MHz Intel Pentium 4 2408 MHz

FSB 150 MHz 150 MHz

Memory 512 MB 512 MB



Display Information

Graphics Chipset ATI RADEON 9700/9500 Series ATI RADEON 9700/9500 Series

Driver Name RADEON 9800 PRO RADEON 9800 PRO

Driver Version 6.14.1.6307 6.14.1.6307

Video Memory 128 MB 128 MB

Core Clock 302 MHz 302 MHz

Memory Clock 302 MHz 302 MHz



Sound Information

Sound Adapter Driver Name SB Audigy Audio [A000] SB Audigy Audio [A000]

Sound Adapter Driver Version 5.12.1.253 5.12.1.253



Benchmark Settings

Program Version 3DMark03 Revision 1 Build 3 3DMark03 Revision 1 Build 3

Resolution 1024x768@32 bit 1024x768@32 bit

Texture Filtering Optimal Optimal

Pixel Processing None None

Vertex Shaders Optimal Optimal



Main Test Results
3DMark Score
(3DMarks) 4623

4623


CPU Score
(CPUMarks) 491.0

491.0




Game Tests

GT1 - Wings of Fury
(fps) 162.2

65.1


GT2 - Battle of Proxycon
(fps) 30.0

6.7


GT3 - Troll's Lair
(fps) 27.0

6.6


GT4 - Mother Nature
(fps) 27.4

13.7

[GlowStick]

Originally posted by 5150 Joker
Wow then you must be a complete moron that cannot build a computer properly. I can confirm his score with my own: http://service.futuremark.com/compare?2k3=655196

R9700 using 3.2 cats modified to 9800. Core/mem set at 301/301 using R3D Tweak. No wonder you get stuttering and other issues, it's obvious your rig has problems due to a severe PEBKAC issue. :rolleyes:

LOL

That is all you retard ATi'ers ever can say "Dude, you must of forgot to put your Video card IN your slot, I DID THAT TOO, then i realzied you have to install hardware, but it took me a long time to figure that out"

You really are retards. It dose make me giggle.

For fairness, i will test the videa card on 2 of my friends rig

One of them is a 2200+ on a VIA KT333 chipset, and the other has a silimar setup with a 2000+ but Watercooled.

[5150 Joker]

Originally posted by GlowStick
LOL

That is all you retard ATi'ers ever can say "Dude, you must of forgot to put your Video card IN your slot, I DID THAT TOO, then i realzied you have to install hardware, but it took me a long time to figure that out"

You really are retards. It dose make me giggle.

For fairness, i will test the videa card on 2 of my friends rig

One of them is a 2200+ on a VIA KT333 chipset, and the other has a silimar setup with a 2000+ but Watercooled.

You just got owned, be a man and admit it. Your score is crap because your rig sucks and was obviously built by you. I'm not an fanATIc, in fact I'm very impartial and have never owned any ATi product prior to this R9700 Pro. My next card is still undecided depending on how well NV35 performs vs ATi's R350 with 400 core and DDR II memory. :rolleyes:

P.S. if your friend's rigs are built like yours, we'll probably see a repeat of your abysmal results.

[GlowStick]

Originally posted by 5150 Joker
You just got owned, be a man and admit it. Your score is crap because your rig sucks and was obviously built by you. I'm not an fanATIc, in fact I'm very impartial and have never owned any ATi product prior to this R9700 Pro. My next card is still undecided depending on how well NV35 performs vs ATi's R350 with 400 core and DDR II memory. :rolleyes:

P.S. if your friend's rigs are built like yours, we'll probably see a repeat of your abysmal results.

LOL that just shows how much you dont know.

Because when my friends rigs show the exact same results all you can say is "NUH HUH, they built their rig wrong".

Now, after looking though the Database from Futuremark.com i am seeing my results for the stock clock speeds of the 9800pro are *ta da* amazeingly close! wow within 100points. Oviously they built their rig wrong also.


Now lets see, explane how you built your rig diffrently than mine!

Did you do a magical dance while you put your processor in the socket? Do you have an secret of slideing a video card into a slot? Oviously you do.

Or could it be that your upped fsb, 'modded' drivers and mods had anything to do with this?

[5150 Joker]

Originally posted by GlowStick
LOL that just shows how much you dont know.

Because when my friends rigs show the exact same results all you can say is "NUH HUH, they built their rig wrong".

Now, after looking though the Database from Futuremark.com i am seeing my results for the stock clock speeds of the 9800pro are *ta da* amazeingly close! wow within 100points. Oviously they built their rig wrong also.


Now lets see, explane how you built your rig diffrently than mine!

Did you do a magical dance while you put your processor in the socket? Do you have an secret of slideing a video card into a slot? Oviously you do.

Or could it be that your upped fsb, 'modded' drivers and mods had anything to do with this?

A higher FSB will play a small role in a video dependent benchmark like 3DMark03. If i had a 2.53 ghz cpu lying around, I'd be glad to run it at default fsb and prove you a moron once again. I'm sure someone else around here has one that will be glad to do so. It's amazing, the evidence is right in front of you yet you aren't willing to accept it. A slightly higher fsb will NOT yield more than twice the score you got. Some people just don't know wtf they're doing, you seem to be one of them. I can't tell you what is wrong with your system, I'm not there to look at it nor do I care. I just couldn't resist posting my score to prove you an idiot for jumping to conclusions and labeling the other poster a liar.

[Rogozhin]

Glowstick

5150 is correct. Your system is FUBARED big time if you only get 1500 marks.

I am running xp 2100+ and a soyo KT333 Lite and I am getting 4600.

Here is the url

http://service.futuremark.com/compare?2k3=579838

Might need to go into another hobby and let pros build your systems.

Rogo

[5150 Joker]

Originally posted by Rogozhin
Glowstick

5150 is correct. Your system is FUBARED big time if you only get 1500 marks.

I am running xp 2100+ and a soyo KT333 Lite and I am getting 4600.

Here is the url

http://service.futuremark.com/compare?2k3=579838

Might need to go into another hobby and let pros build your systems.

Rogo

Rogozhin,

If you have access to R3D Tweak, set your clocks to 301/301 and run the benchmark please. I'm willing to bet your score won't drop below 4k.

[Rogozhin]

Also

Most of your conclusions regarding your posts are based on fallacious premises, same as marco.

I am too tired to drub your sorry posts right now. So i will have to say goodbye and look for me later.

Rogo

[ugapug]

Originally posted by 5150 Joker
Unreal II is a cpu hog though. It runs like a dog on most systems no matter what card you have.

I don't know about your system, but with mine Unreal II runs like butter. 2xAA 4xAF 1280X1024 all settings maxed.
Specs:
Abit NF7-S rev 1.0
T-Bred 1700@190x11 (2112)
TwinX512 3200LL
PNY GeforceFX 5800 Ultra
WD Raptor SATA
WD 1000BB 100gig (w/2meg cache)

Upgraded from a Geforce4 Ti4600. Had an ATI 9700 Pro to compare to the Ultra. IQ seems quite similiar to me, but the colors on my Ultra with Digital Vibrance activated seem a bit more vivid.
Just my .02
And I like loud fans. It gives me an excuse to crank up the speakers :D