New Nvidia Drivers Results...

[Razorscott2YK]

If you have an Nvidia card and the new drivers and want to post your 3DMark results or just wether they worked or not, feel free to post!

Those of you who have questions about the drivers... feel free to poat as well!

-Razor2YK

 

[Xtasy0]

3dmark is meaningless, why not goto a site like hardocp.com and read their preview and look at their benchmarks which show you what kind of experience you'll get in GAMES, you know, the thing you bought the beefy videocard to Play, GAMES, unless you just sit and pretend you're playing 3dmark.... ;/

 

[Washuu]

true, i got 1887 on 3dmark03, instead of 1942.....
but my fps increased 10-15, on unreal 2 xmp.
so that made me happy, i have a dx8 card, and it now runs dx8 games better, but it had an adverse affect to dx9 performance, so, meh, ill do a TRUE 3dmark01 test for dx8, and see what i get there

 

[Washuu]

nope, lower 3dmark01 test too, my hypthosis was incorrect, who knows, maybe nvidia has stop 3dmark test cheating
hahahaha.
but yeah, im games, its doing good for me

 

[Xtasy0]

Originally posted by Washuu
true, i got 1887 on 3dmark03, instead of 1942.....
but my fps increased 10-15, on unreal 2 xmp.
so that made me happy, i have a dx8 card, and it now runs dx8 games better, but it had an adverse affect to dx9 performance, so, meh, ill do a TRUE 3dmark01 test for dx8, and see what i get there

how do you measure a cards dx9 performance, when the card cannot perform dx9 functions?

 

[NSPIRE]

Good question... I was wonderin' the same damn thing, lmao.

 

[TheRook]

Originally posted by Xtasy0
how do you measure a cards dx9 performance, when the card cannot perform dx9 functions?

all it does is run the benchmarks it can... but without the directx9 shaders/technology so it does the wings of fury demo and i think the action demo too...

TheRook

 

[Arno]

Originally posted by TheRook
all it does is run the benchmarks it can... but without the directx9 shaders/technology so it does the wings of fury demo and i think the action demo too...

It's obvious that Xtasy0 and NSPIRE were kidding, but I'm not sure if you got the joke.

FYI, GeForceFX cards are indeed equiped with DX9 tech and they are capable of running for example the Mother Nature demo. Although with a bad framerate.

 

[Cyphron]

new drivers help a little in the fps department, but i actuallylost 3dmark points in 01 SE and 03. i havent run aquamark with these yet. still going to enjoy my 9800 pro. this 5200 is total crap and needs to be shot when i put my new card in.

 

[crabcakes66]

"equipped" is kind of subjective........considering that they dont give playable FPS in alot of cases without using ultra-optimised code and NOTICABLE image quality loss......

thats not really "equipped" if you ask me....... i can say my 92' ford tuarus is "equipped" with a turbo............but if it still takes 30 minutes in the quarter.....then you can take a wild guess and figure out that its not.

 

[TheRook]

Originally posted by Arno
It's obvious that Xtasy0 and NSPIRE were kidding, but I'm not sure if you got the joke.

FYI, GeForceFX cards are indeed equiped with DX9 tech and they are capable of running for example the Mother Nature demo. Although with a bad framerate.

may have been a joke may not... but text is difficult to define sarcasm... when done badly...

i never mentioned the fx range as directx8...

Originally posted by Xtasy0
how do you measure a cards dx9 performance, when the card cannot perform dx9 functions?

this is the question... and as somebody else asked i figure id answer...

and "FYI" please dont try and make out i dont know what im talking about if you cant read what my post was referring to dont quote me... and make an ass outta yourself!

Regards

TheRook
(knows more useless crap and information than you think...)

 

[Xtasy0]

i wasn't kidding, did you see who i quoted? the person i quoted has a geforce4 and said the new drivers adversely affected his dx9 performance, i was just asking how he could measure his dx9 performance on a geforce4.

 

[Arno]

Sorry, my mistake. I misread the quote.

 

[NSPIRE]

Originally posted by Xtasy0
i wasn't kidding, did you see who i quoted? the person i quoted has a geforce4 and said the new drivers adversely affected his dx9 performance, i was just asking how he could measure his dx9 performance on a geforce4.

Same here. I didn't mean to sound (err... look in this case) sarcastic / rude. I was bein' serious about it, though.

 

[Ridic]

Originally posted by crabcakes66
"equipped" is kind of subjective........considering that they dont give playable FPS in alot of cases without using ultra-optimised code and NOTICABLE image quality loss......

thats not really "equipped" if you ask me....... i can say my 92' ford tuarus is "equipped" with a turbo............but if it still takes 30 minutes in the quarter.....then you can take a wild guess and figure out that its not.

...and im sick of all the nvidia dudes (not so much on these forums..) telling me there is no IQ loss...

 

[TheRook]

Originally posted by Ridic
...and im sick of all the nvidia dudes (not so much on these forums..) telling me there is no IQ loss...

in previous versions of there drivers... ill admit they have had really poor image quality... (and i really love nvidias cards when they do the job properly)

but the latest release of drivers... the image quality is matching ati's so there is no more going well nvidias cards have bad image quality...

you can just say nvidias cards cant use pixel shaders for shit...

TheRook

 

[CokeLite]

Before update Low 2700's
After update high 2700's

Using 3dmark01

BTW I am using an overclocked 5200FX 128

 

[TheRook]

are you sure ur getting THAT score on 3dmark 2001??

jesus christ if you are something is seriously wrong...

even tho its a 5200 its still appaulling!

return it if you can or something...

whats the rest of your specs...

TheRook

 

[Razorscott2YK]

My results...

Im not sure if this is a good result or not, coz this is the first time ive used 3dmark on my new system, but here are my results.

3dMark2001 with 45.23 drivers: 6895
3dMark2001 with 52.16 drivers: 7970

I got a pretty good increase with these drivers, not to mention the image quality in games like Call Of Duty and C&C Generals was improved.

My total benchmark for AquaMark2003 was: 14325

My system is:

Asus P4s533-x Mainboard
Geforce4 Ti 4200 128mb
P4 2.2ghz
512mb pc2700 ddr ram

And ive had no problems with games using the Quake 3 engine, like some people have reported.

just thought id put in my two cents...

Razor2YK.

 

[TheRook]

your scores also seem low...

your running this on default setting? 1024x768 32bit 0xAA etc?

my friend has a similar spec pc to you RazorScott2YK

and his score is near 9-10 thousand...

but you spec seems high enough... so i dunno whats wrong...

try defragging your hddisk

but im not sure why your score is so low... should be higher!

TheRook

 

[CokeLite]

I am running a 2.4 P4 with 512 MB RAM and the 5200FX on a 645 mobo...

I meant 4700's, not 2700's My bad

 

[Shuzer]

Originally posted by CokeLite
I am running a 2.4 P4 with 512 MB RAM and the 5200FX on a 645 mobo...

I meant 4700's, not 2700's My bad

And that's 3dmark01?!

 

[CokeLite]

I dunno WTF should I do get a 9800 pro or something?

Gonna be kinda hard right now cause I just bought a HDTV for my place.

 

[Shuzer]

Originally posted by CokeLite
I dunno WTF should I do get a 9800 pro or something?

Gonna be kinda hard right now cause I just bought a HDTV for my place.

Anything would be better.. (well, not anything, but you get the picture)

For lower-budget cards, I'd suggest..

If you want to go nVidia, only card I'd go for would be a GeForce FX 5700 ultra, it's pretty good I hear (for the price, performs better than the 9600XT in some cases, but it's not out yet.. although right now I'd still prefer an ATI)

Radeon 9600XT or Radeon 9600 Pro

Or if you have a little more money to spend, a 9700/pro

 

[CokeLite]

OK. I am comparing my results to other peoples right now and am getting pretty pissed off... WTF a guy with the same exact specs is in the 8000's

 

[Shuzer]

Do you have anti-aliasing or anisotropic filtering on? If so, you should disable those.. that'll be a big performance hit (especially for that card)

 

[CokeLite]

Yeah its disabled

 

[ShaithEatery]

Originally posted by Ridic
...and im sick of all the nvidia dudes (not so much on these forums..) telling me there is no IQ loss...

hey, how bout reading the review on tomshardware.com... looks like there's a little controvercial "optomizations" on those fancy pancy new catalists while nvidia made this strict guideline on the new driver's development.


http://www.tomshardware.com/graphic/20031023/nvidia-nv38-nv36-13.html

http://www.tomshardware.com/graphic...ml#questionable_optimizations_in_atis_drivers

 

[Shuzer]

Originally posted by ShaithEatery
hey, how bout reading the review on tomshardware.com...

The only problem being, Tom's Hardware is a bit bias towards nVidia (just a bit... :cheese

 

[ShaithEatery]

the proof is in the pics, all i have to say

and wasn't those new catalists burning up ati cards? read it one some topic in here.

 

[Arno]

Originally posted by ShaithEatery
hey, how bout reading the review on tomshardware.com... looks like there's a little controvercial "optomizations" on those fancy pancy new catalists while nvidia made this strict guideline on the new driver's development.

Nvidia's guideline could still be somewhat stricter, but there's a definitive improvement over the 45.23 drivers. The lighting in AquaMark3 looks much better and I was impressed with the quality of the ogre lair scene in 3DMark03. The skin on the ogre looks much more alive now. Apparently Nvidia's new shader compiler pays off, although I still lost some performance points.

System: P4 2.6C; 1GB RAM; FX5600

3DMark01SE 45.23 : 10165
3DMark01SE 52.16 : 10066

3DMark03 45.23 : 2788
3DMark03 52.16 : 2666

AquaMark3 45.23 : 18.738
AquaMark3 52.16 : 17.802

 

[dawdler]

Originally posted by ShaithEatery
hey, how bout reading the review on tomshardware.com... looks like there's a little controvercial "optomizations" on those fancy pancy new catalists while nvidia made this strict guideline on the new driver's development.


http://www.tomshardware.com/graphic/20031023/nvidia-nv38-nv36-13.html

http://www.tomshardware.com/graphic...ml#questionable_optimizations_in_atis_drivers

http://www.elitebastards.com/page.php?pageid=2397&head=1&comments=1.

Thats the response on Toms accusations. But no one, not Toms or EB or you or even me, knows that 'Nvidia made this strict guideline'. Why? Because Toms just threw it out without proving it. He could EASILY have added a DX9 rasterizer picture. That would show which was closest to 'correct' rendering. It wouldnt surprise me if he did the test and was surprised that ATI was closest, but 'forgot' to include it :dozey:
Toms Hardware is one of the worst review sites around. Big yes, but not very good, same as HardOcp.

 

[Arno]

I agree, it's quite obvious that Tomshardware is biased towards Nvidia. HardOCP also briefly mentioned the Nvidia event, but they were extremely negative about Nvida. Not to mention they conveniently 'forgot' to mention the accusations towards ATI. Bottomline is that it's hard to find an unbiased source on the internet.

Right now I prefer AnandTech.com as a source of information. They got very extensive reviews and always try to look at issues from different viewpoints.

Do you have a favorite review site, Dawdler?

 

[Xtasy0]

Originally posted by Arno
I agree, it's quite obvious that Tomshardware is biased towards Nvidia. HardOCP also briefly mentioned the Nvidia event, but they were extremely negative about Nvida. Not to mention they conveniently 'forgot' to mention the accusations towards ATI. Bottomline is that it's hard to find an unbiased source on the internet.

Right now I prefer AnandTech.com as a source of information. They got very extensive reviews and always try to look at issues from different viewpoints.

Do you have a favorite review site, Dawdler?

hardocp didnt attend the nvidia event, therefore they wouldnt have known about the"optimizations" that were mentioned at the event.

and hardocp has a right to be negative to nvidia and their propaganda, nvidia has been shitting all over the end users and lying to hardocp about fixing the anisotropic filtering issue (i don't even know if they've fixed it yet). point is, yes kyle at hardocp tells it like he sees it, which may come out sounding bias, but the reviews of hardware aren't biased, their new preview of the 5950 was congratulated by like the whole community on their forum, everyone is digging their new review style and they all agree they did an amazing job on the 5950 preview.

 

[rafiki]

i got 10,312 on '01 with a
athlon xp 1800@140fsb(1609Mhz)
soyo dragon kt333 MB
Gainwaird Ti 4200(det 4403)
512 of corsair XMS
and DX9
all the settings in 3dmark were at default.
HERE
is the results page.
gonna clean house and try the new det's this weekend.

 

[crabcakes66]

looking at a few more reviews reviews...the 5700ultra looks like a very nice upgrade over the 5600....and is even faster than a 9600Xt alot of the time ...seems like some real improvments were made there.......

it can actaully compete with the 9600xt in the mid-range market

i dont get the 5950 though....that thing is a joke......

 

[Shuzer]

Originally posted by crabcakes66
looking at a few more reviews reviews...the 5700ultra looks like a very nice upgrade over the 5600....and is even faster than a 9600Xt alot of the time ...seems like some real improvments were made there.......

it can actaully compete with the 9600xt in the mid-range market

i dont get the 5950 though....that thing is a joke......

Any idea what the projected cost will be, crab? My brother wants a graphics card for his birthday next month, I was going to see about a 9600XT, but I'm kinda interested in the 5700 ultra

 

[crabcakes66]

around 200$(US)

 

[Arno]

Originally posted by Xtasy0
hardocp didnt attend the nvidia event, therefore they wouldnt have known about the"optimizations" that were mentioned at the event.

The fact that they refused to go is their own fault. Instead they copied the information from a Gamespot article. Gamespot's article does mention the ATI "optimizations", so hardocp should know about it. But they decided not to share this knowledge with their readers.

Originally posted by Xtasy0
and hardocp has a right to be negative to nvidia and their propaganda, nvidia has been shitting all over the end users and lying to hardocp about fixing the anisotropic filtering issue (i don't even know if they've fixed it yet). point is, yes kyle at hardocp tells it like he sees it, which may come out sounding bias, but the reviews of hardware aren't biased, their new preview of the 5950 was congratulated by like the whole community on their forum, everyone is digging their new review style and they all agree they did an amazing job on the 5950 preview.

I've just read the preview. It indeed has the right style. Their conclusion that the 9800XT is a better card then the 5950 is correct and sufficiently supported by the benchmarks. The only gripe I have with the preview is the lack of an OpenGL benchmark (which the FX card would probably win). I understand that it's just a preview, but quite a few important game titles are in OpenGL. Now and in the future.