Holy CRAP......NV40's 3dMark03 Score = 12,535!

Originally posted by LabRat
HAHAHAHAHAHAHAHAHAH!!!!

Wait, you were joking, right?

The Inquirer is close to right most of the time..

just because once or twice in a long time they get no where close, but usually they are close if not on the dot.

just look at what they had to say about Prescott... no one believed them, and they were right!

please, leave the inquirer bashing aside.
 
Originally posted by Stang Man
The Inquirer is close to right most of the time..

just because once or twice in a long time they get no where close, but usually they are close if not on the dot.

just look at what they had to say about Prescott... no one believed them, and they were right!

please, leave the inquirer bashing aside.


Why should we leave them alone? I don't trust their info. they come out with so called "leaks" that no one else seems to have and I find that hard to believe. Plus it's our opinion and my opinion is that they are not a 100% reliable source.
 
It's a benchmark, it's pretty easy to play up to a benchmark

Nevermind the fact is the inquirer.
 
I won't be able to afford either card, so... I just get to laugh at all the people arguing and speculating :D
 
Well, I suppose you could say the NV30 is DX9, but it wasn't really.

It was superior in that it exceeded the spec (32FP) and also was inferior to it (16FP) I would never say it actually met the spec though.

Its sort of like cars, sure you can have a part that runs only at 2000 RPM and only at 4000RPM, but if it doesn't run at spec it cause more problems that its worth. 4000 RPM might burn out some other components, 2000 might be inadequate for some parts to function properly, where the spec 3000RPM part will work just fine. Its especially important on things that cannot be adjusted easily.

And in the case of videocards, it obviously cannot be adjusted easily if they have to make completely different renedering engines for different types of videocards.
 
Originally posted by ZenOps
Well, I suppose you could say the NV30 is DX9, but it wasn't really.

It was superior in that it exceeded the spec (32FP) and also was inferior to it (16FP) I would never say it actually met the spec though.

Its sort of like cars, sure you can have a part that runs only at 2000 RPM and only at 4000RPM, but if it doesn't run at spec it cause more problems that its worth. 4000 RPM might burn out some other components, 2000 might be inadequate for some parts to function properly, where the spec 3000RPM part will work just fine. Its especially important on things that cannot be adjusted easily.

And in the case of videocards, it obviously cannot be adjusted easily if they have to make completely different renedering engines for different types of videocards.

as long as you don't put the partial precision hints in the game the card will run at FP32 by default, and the R3xx will run at FP24 by default

technically it is over the spec therefore valid, no matter what your opinion is on the matter, those are the facts
 
True, it doesn't really make that much of a difference with FP. But there is a difference and gamemakers are using card-specific optmizations so its there for sure.

Techincally FP32 is better quality than FP24. Nvidia seems to like to call it DX9+ and seems to mysteriously avoid calling it something like DX9 and DX9+ compatible.

It sort of reminds me of when ATi trying to do 5-6-5-0 RGBA with seperated alpha functions, technically I did not think it was in spec for DX at that time either. 5-5-5-1 was supposed to be the standard, but it was assumed that ATI's alpha engine was too slow, so they sort of fudged the specs a bit and used a specialized handler...

There is always room for stuff like that, and I would not be surprised to see Nvidia trying the Hybrid 16-FP on faraway objects and 32-FP on closeup objects. I'm not saying its wrong, its just not to spec.
 
Originally posted by VIOLENCE FACTOR
...........please............kill...............Inquirer.............God........please........ :(

Originally posted by IceWind
This article is good...... for me to poop on!!!

:D :D :D

thats about all that needs to be said

i remember when nv35 was gonna get >22k on 3dmark2001 stock.. yeah that almost happened
 
BTW: There is no standard on colour compression either (At least normal texture compression is relegated to some form of standard, S3TC DXTC1-3)

I have an odd feeling that NVidia has been overcompressing the colour values and then recompensating with digital vibrance in the drivers of late... Its starting to become noticable to me (me being a colour freak)
 
I have an odd feeling that NVidia has been overcompressing the colour values and then recompensating with digital vibrance in the drivers of late... Its starting to become noticable to me (me being a colour freak)
I noticed this as well, however, I don't know if they are doing it because they have to or because they want to - I kind of like the effect- because the LCD screens are naturally too bright for my eyes so the duller textures make it better for me, and then, with a CRT screen, I compensate easily with the Digital Vibrance control and get as much 'brightness' to the colors as I want.

It's a particularly nice feature of their drivers IMO.
 
The inq is all about going on tiny leads and creating big stirs (this one for example). yes, they are sometimes wrong... very wrong... but now and then they provide accurate info. they get tons of publicity by putting up really crazy stories that everyone talks about. everyone reading this thread that didnt know what the inquirer was does now. its just what they want. anyway, about that 3dmark score... '01 was pretty good for benchmarking and comparing but I think we all know just how flawed '03 could be with drivers and optimizations. im waiting for real performance tests and numbers. we have yet to see any hard proof of which chip will wear the crown.
 
Do you all really think that that score is unrealistic? If we can get a top of the line system right now getting 7382 3dMarks (with the 9800XT), 12,535 isn't that much more, especially when you figure in the entirely new architecture and that these cards have been worked on for a long time and are next generation products. I agree that we can't judge this by 3dMarks alone, as they can be hacked, but I am pretty sure that nVidia will NOT cheat again, not after the reactions of the enthusiast community the first time around. Those of us who haven't been turned into ATI fanboys and still have hope for nVidia ( a small amount I know ) would never buy nVidia again. They aren't going to risk it.

Is the Inquirer telling the truth? We have no idea, but this score is realistic, and with NV40 parts not so far away it is possible that this is leaked information.

We can't judge companies and their products by the past, barely any companies have a record that's 100% clean. Wait until the benchmarks come out, then you can call the NV40 a horrible product if it can't perform. Until then, it would be stupid to assume that it's gonna be another GeForceFX like failure.
 
Originally posted by Spinal
Why should we leave them alone? I don't trust their info. they come out with so called "leaks" that no one else seems to have and I find that hard to believe. Plus it's our opinion and my opinion is that they are not a 100% reliable source.

It's called journalism.

The Inquirer is like the ditzy little blonde girl you had back in high school that always gossiped through English.

Even though you *tried* not to listen in, and even though you *knew* you shouldn't give a second thought to what she said, you kept one ear open and you enjoyed every second of it

Half the time she turned out to be right, too, even though she always had some kind of opinionated spin on the story.

There are all kinds of resources on the web for factual information as well as rumour mongering - none of them should be censored, no matter how much it might irritate the fanboys. You don't like the Inquirer? That's nice... whatever flips your pancake.

As far as their "leaks" that "no one else has"? There are LOTS of places that have inside information - nothing special. As long as there are human beings working in a particular field or industry, there will be a steady flow of outward information no matter what - it's human nature. Some of it may become twisted or fabricated in the process, and some of it may even be pure misdirection from the manufacturer themselves; there is still plenty to look at provided you approach the information with an open and analytical mind.

Long story short, for those of you who skipped through the post because it was "too long and made my brain hurt" - INQUIRER OWNS JOO HOOHAHAHA HOO HOO!!11 :rolleyes:
 
Half the time she turned out to be right, too, even though she always had some kind of opinionated spin on the story.

Maybe the analogy would be apt if you said:
"You think she was right half the time because that's what she told you, but since you weren't paying much attention, you can't be sure, and even though you recorded every conversation with her, you cannot be bothered to actually check how often she was right because that would take time and energy."

Tabloid journalism is not the definition of journalism.
 
Originally posted by leukotriene
Maybe the analogy would be apt if you said:
"You think she was right half the time because that's what she told you, but since you weren't paying much attention, you can't be sure, and even though you recorded every conversation with her, you cannot be bothered to actually check how often she was right because that would take time and energy."

Tabloid journalism is not the definition of journalism.

Sorry, but I'm the anal jackass who checks up on everything ;)

I can't say for sure what the Inquirer's success rate is, as I probably read less than 1/3 of their stories, but I find them at least entertaining.

Certainly no one should use them as a factual source of discussion - no more than support material.
 
They were 100% correct about Prescott, giving relatively concrete info while all the 'respectable' news sites were bound by NDAs.

The ATI damage control guys will likely be eating their words about this time two weeks from now.
 
but you guys always bring up how "unreliable" the Inq are... when in fact, they are probably at least 75% of the time right. that's huge for what they get their info from.

and who cares, if they aren't right, it was all in good fun.

i'm not gonna cry if NV40 comes out and it makes 11,3k in 03...:rolleyes:
 
However far-fetched The Inquirer usually sounds, they usually end up being at least partially right.
 
Because of this thread i loaded the inq archives other than a few things i saw (call the xt the 9900) they were, i would say 90% or so correct on what they said.

Browse their archives. (i just looked at the video card section, and no thats not a scientific number hehe hehe)
 
I hate ATi. I don't like nVidia either, but i definetly hate ATi more. I've had a couple bad experiences, first with a Radeon 7000 and then a 7200; both were 64 meg. Neither worked correctly, or nearly as well as the 16 meg Riva TnT i was trying to upgrade from lol! I was getting completely fucked up displaying of lightning, textures and particles in OpenGL; HUGE drop in performace on anything 3D; big gain in performace on anything 2D though. Bah! Had to get a geforce instead. :mad: :mad:
 
Which ever card runs faster with better quality gets my money.

If they were 100% equal then I would go for Nvidia because I can't stand ATI's dual monitor implimentation or their shitty drivers. Some older games especially won't even run on ATI based cards at all. Or they do funny things.

Before anyone calls me an Nvidia fanboy check the sig. I own an ATI card. I just don't like ATI software. But currently their hardware is king of performance.
 
I can't believe there are people who have already decided what card they are gonna get :rolleyes:
 
Originally posted by AcneBrain
I can't believe there are people who have already decided what card they are gonna get :rolleyes:

I really doubt most people have made up their minds, they may just be starting to get a preference for which card they want. But they may change their preferences later on, since you can't actually buy a card right now, and unless you want to live under a rock, you'll be forced to see some benchmarks for both cards. Thus, what they're thinking now will be irrelevant when they have a chance to make an informed decision.
 
It's funny because that gets more in 3D Mark 2003, than I get in 3D Mark 2001. :D
 
Originally posted by Mojo
It's funny because that gets more in 3D Mark 2003, than I get in 3D Mark 2001. :D

My measly GF4 4400 gets 1890 in 3Dmark2003, im soooo looking forward to upgrading. Hmmmm the thought of playing ut2004 at 1600x1200 makes me drool :cool:
 
Hmmm....

Originally posted by onetwo
They were 100% correct about Prescott, giving relatively concrete info while all the 'respectable' news sites were bound by NDAs.

The ATI damage control guys will likely be eating their words about this time two weeks from now.

:p
 
lol funny somebody brings this thread back from the dead...reading all those people on the first couple of pages just makes me laugh.

rofl
owned.
 
Originally posted by Bad_Boy
lol funny somebody brings this thread back from the dead...reading all those people on the first couple of pages just makes me laugh.

rofl
owned.
 
IMO this clearly shows how much 12k in '03 really means. This guy in the link in the post above had his machine cranked to the brink and then some. The 6800U can outscore even him at default clocks.
 
R420 Scores around 15,000 in 3dmark03...

But by all means dont take my word on it ;)
 
Originally posted by Sir-Fragalot
Which ever card runs faster with better quality gets my money.

If they were 100% equal then I would go for Nvidia because I can't stand ATI's dual monitor implimentation or their shitty drivers. Some older games especially won't even run on ATI based cards at all. Or they do funny things.

Before anyone calls me an Nvidia fanboy check the sig. I own an ATI card. I just don't like ATI software. But currently their hardware is king of performance.
That is complete and total FUD...

First Ati's Dual monitor Solution is far better than Nvidias and has been since its inception. Hydravision rocks.

Secondly The utter garbage you are spwing about their drivers is just that utter GARBAGE that is completely untrue..

Thirdly.. I have never run accross an older game that would not run on Any modern ATi card.

You can claim you "own and Ati card" but if you really did imo you would not be posting untrue trash like the above. Especially a 9600pro. Which i am using im my laptop now and have not one single issue that i can think of on any of the games i have on it. Hell it even plays the origional Battlecruiser just fine.. :rolleyes:
 
Originally posted by Swat
I didn't read through this whole thread but I found this

http://service.futuremark.com/compare?2k3=2113980

If you look at that overclocked XT and are a "bench junkie" 12k aint too impressive IMO.

Ummm....ya. 12K is at stock speeds. That XT also has the CPU at 4500mhz. If you put a OCed 6800U in that system with that CPU there is little doubt it would be at least 25-35% faster. A score of 12k despite all the previous scandals is impressive non the least.
 
That is complete and total FUD...

First Ati's Dual monitor Solution is far better than Nvidias and has been since its inception. Hydravision rocks.

Secondly The utter garbage you are spwing about their drivers is just that utter GARBAGE that is completely untrue..

Thirdly.. I have never run accross an older game that would not run on Any modern ATi card.

You can claim you "own and Ati card" but if you really did imo you would not be posting untrue trash like the above. Especially a 9600pro. Which i am using im my laptop now and have not one single issue that i can think of on any of the games i have on it. Hell it even plays the origional Battlecruiser just fine..

Hellbinder? From Rage3d? Well, hello! You probably don't recall me, I only have about 2000 posts there. 'XanderF'.

Anyway, I think you are over-reacting.

First off, ATI's multi-monitor support DOES suck. Seriously, have you actually TRIED nView? You don't even need to use Ultramon with it, like you do with ATI cards. Trust me, nView *is* better than Hydravision.

Secondly, yeah, their drivers are inferior. Some older games are fine, some are not. 16-bit FSAA, anyone? It's not even a problem with nVidia cards, but ATI is really spotty on it. ATI's FSAA sure *looks* a LOT better, when it does work - that is changing with the 6800, though. nVidia will at least be competitive in that area again. Stereoscopic 3d support? nVidia has it, ATI doesn't. Application profiles? nVidia has it, ATI doesn't. Fixed-aspect ratio scaling on LCDs? (Say you want to play your 640x480 game at 1280x960 'letterboxed' on a 1280x1024 LCD) Again, nVidia has it, ATI doesn't. Want to set refresh rates individually for each resolution your monitor supports? nVidia can! ATI needs a third-party util to. I could go on and on and on.

And let's not even talk about who has an easier to use control panel!

ATI drivers have improved VASTLY. They really have. They are not currently unreliable in any way. They don't crash, I would not call them buggy.

But, after having used both a LOT - and recently - I can say with some authority that nVidia's drivers just feel more POLISHED. And that counts for a lot.

I do agree with your criticism of the original posters "some games don't work with ATI" - that's not true, anymore. At least, I've not seen a post on ATI's forums in a VERY long time (?years?) about a game that just wouldn't run, period, on ATI hardware. ATI's latest drivers really are heads and tails above where they used to be.
 
lol most of you people are fools...you act like your fuckn your ati or nvidia graphics card and can never trade it in...wake you fools its only a card not your wife. geez just be happy ati and nvidia are comeing out with new cool cards.... you people amaze me.... what nerds.
 
*I* was attemting to post a balanced opinion.

What graphics cards have I owned to form said opinion?

3dfx Voodoo3 -> 3dfx Voodoo5 -> ATI Radeon 64 DDR -> GeForce3 Ti200 -> GeForce3 Ti500 -> GeForce4 Ti4200 -> ATI Radeon 9500 Pro -> ATI Radeon 9800 Pro -> ATI Radeon 9600 Pro -> ATI Radeon 9800 non-pro -> GeForceFX 5900XT

As you can see, the majority of my 3d accels were ATI cards, with an almost balance of nVidia cards, to boot. I have no brand loyalty whatsoever, I use whatever has the best price/performance/package value at the moment I get a new card.

My observations are strictly objective.
 
Having used both nVidia and ATI (Current setup is in my sig but theres been plenty others) I have to say I disagree.

Saying you like one set of drivers is like saying you like Windows or Mac, its personal preference. I actually like the ATI drivers better. Also, having dealt with dual monitors on both systems with jsut about every computer, I like ATIs better, it handles different resolutions across the screens better than nVidia does I find.

As far as support for old games, I dunno. Not an issue for me.

I won't personally say either has truly superior drivers, because I just haven't seen that as being the case over the past couple years.
 
Back
Top