NV40 benches (in case u didnt see)

[CaiNaM]

it's not about being a fanboi of anyone..

rollo simply stated the games and the resolutions he plays at, his 9700p was no faster than his 5800u. it's true, so why turn it into a big deal? while he might have "passionately" argued his point, it was a reactive one, not proactive. and frankly, if you didn't spend too much time getting caught up in who was right and who was wrong (it's like one was 50% right and the other was 60% right on much of that stuff), it was quite funny.

as for the price.. bummer

you won't see me spending $499 for a marginal improvement over my 9800p or fx5900.... but again, it's a bit too early to tell.

 

[nRollo]

Originally posted by: CaiNaM
actually he doesn't troll at all..

at any rate, yea, i see your point about "wanting more", but it's still early, we don't have any comments so we don't know what shape drivers are in, or what the context of the test were, or what kind of cpu's were being compared.

i would also like to see more, but i'm not going to make too much out of it at this early point.

the other thing to consider also is price. the 9800xt is the more expensive card if the 6800u is indeed $399... lots of unanswered questions yet.

Sniff...sob...he caught me fair and square Cainam. I buy the cards by all companies just to anger BFG, because it's worth hundreds of dollars to me to confuse him.

I'm always trying to say both brands have their strengths, but for the most part are equal, because I know that will enrage everyone.

 

[nRollo]

I guess we are just fanboys of different forum members.

Yikes. These boards would be sad indeed if guys like me and BFG had "fanboys".

Carmack deserves fanboys, BFG and I deserve a read if you think we might have something interesting to say.

As far as the pricing goes, I'm still not disappointed if the only difference in the cards is 50MHz RAM speed. The 6800Pro would be worth $300 to me. (and I'd pay that over $500 for a card with 50MHz more memory speed any day)

I want to see the real reviews and ATIs offering first in any case.

 

[reever]

it's like one was 50% right and the other was 60% right on much of that stuff)

Is there a 3rd party that is -10% right?

 

[CaiNaM]

you know.. it's also possible the benches aren't even authentic.. take some hardocp reviews, use a bit of photoshop..

specs on the 6800u from gzeasy were 400+/1200

the pictures of the driver cp show 400/1100 - 6800nu speed.. the driver cp also doesnt show "6800ultra" as it should.

NVIDIA_NV40.DEV_0040.1 = "NVIDIA GeForce 6800 Ultra"
NVIDIA_NV40.DEV_0041.1 = "NVIDIA GeForce 6800"

perhaps that explains some of the "disappointing performance"? guess we'll know for sure in the morning.

 

[reever]

Originally posted by: CaiNaM
you know.. it's also possible the benches aren't even authentic.. take some hardocp reviews, use a bit of photoshop..

specs on the 6800u from gzeasy were 400+/1200

the pictures of the driver cp show 400/1100 - 6800nu speed.. the driver cp also doesnt show "6800ultra" as it should.

NVIDIA_NV40.DEV_0040.1 = "NVIDIA GeForce 6800 Ultra"
NVIDIA_NV40.DEV_0041.1 = "NVIDIA GeForce 6800"

perhaps that explains some of the "disappointing performance"? guess we'll know for sure in the morning.

But how the heck is somebody supposed to accurately photoshop a histogram?

Did anyone mention what kind of processor was being used?

It's a high-end P4, maybe a prescott

 

[VIAN]

I'm not as impressed as I would hope to be.

Well, let's just say that I really don't think I will be buying that when I could just get a 9800 XT for cheaper. I think they have finally made a worthy competitor to the 9800 XT. Now all they have to worry about is the R420.

I see they heard me and now, we can finally turn off brilinear. If we wanted to. It is a very nice optimization. I like the features. I think it is a nice card but overpriced for my needs at the level of performance it gives over the 5950 and/or 9800 XT. It performs mostly at 10-15fps more at times at AA and AF settings that I personally don't care about. Picture quality is also excellent. Similar to 9800 XT. Sometimes a little better and sometimes a little worse.

 

[CaiNaM]

But how the heck is somebody supposed to accurately photoshop a histogram?

you wouldn't have to; all you have to photoshop is the legends (text).

and again, why does the control panel show non-ultra clockspeeds? or no "6800ultra" in the cp title? i sincerley doubt kyle/brent would benchmark a non-ultra thinking it was an ultra....

doesn't that seem strange to you?

 

[nRollo]

If that is a NU, those are very impressive benches. You can't touch a 9800XT for $299, even refurb, and that card won every bench, most at higher settings.

 

[DarkKnight]

I thought it would be a lot faster. Not gonna upgrade my radeon 9800 anytime soon

 

[CaiNaM]

lmao....

check this out

too funny

 

[VIAN]

This card is 8X2 pipes , i bet my ass on this ,...it will be exactly like with NV30 , full 8 pipes who turned up to be 4x2 .
The card is not impressive , is just an improvement over 5950 ,... it might be better than 9800XT proly with a 20 % average (if you even the aa/af) but still it lacks some serious IQ .
R420 will smash this card .

The card is a 16x1/32x1, specifically pointed out. IQ is amazing. I don't know what you are talking about. IQ was definitely great and better than the 5950.

I don't know what you people are griping at too with the 5800U. It is a great card and can defintely kick ass. And driver improvements did help a lot - although it was a bit much, much later, but... it helped.

Sniff...sob...he caught me fair and square Cainam. I buy the cards by all companies just to anger BFG, because it's worth hundreds of dollars to me to confuse him.

I'm always trying to say both brands have their strengths, but for the most part are equal, because I know that will enrage everyone.

I think the review was done wrong. If they would have made the cards show at the same settings. Then we could get a more accurate display of the 68U's performance. I think more people will be impressed tomorrow when REAL benchmarks are done.

 

[ronnn]

Lets hope it was fake! Would be a good joke if nothing else, as it sure got all the forums going. The alternative would be that gpu's have reached a plateau, much like cpu's. Seems that the rate of progress has slowed, as my system is almost a year old and is not totally out of date. The upside is that keeping current would not be as expensive.

 

[VIAN]

"again, for all you people that think this is a fake"

The benchmarks that took is actually going against them in my opinion. Once they take benches with all the cards at the same settings, then you will cheer.

 

[CaiNaM]

Originally posted by: VIAN
The benchmarks that took is actually going against them in my opinion.

english, please vian.. i don't understand the point you're trying to make...

 

[VIAN]

Ok, I'll try, I'm still in college you know.

I'm saying that (hopefully you'll understand)... that making all the fps similar and showing the different settings doesn't give you an accurate display of performance. And this makes the bechmark negative in favor of the 6800 Ultra. If all cards were benched under the same settings such as COD, we could see more accurately how much of a performance advantage the 6800 Ultra has over the other two cards because we can see the fps. Doing this would have benefited the 6800 Ultra. You notice how people respond to the number difference in COD. And this is a benchmark were ATI is the winner in previous generation benches. -according to VGAIII.

Tomorrow, when benchmarks are done properly, we can aspect COD cases to appear in more benchmarks.

BTW funny link Cainam

 

[BFG10K]

The Far Cry benchmarks are a bit disappointing for sure but there could be a rational explanation for it. I'll wait for the final reviews to see how that issue pans out.

I buy the cards by all companies just to anger BFG,

You anger me to about the same level as a village idiot would (i.e. I'm not angry, I'm laughing).

because it's worth hundreds of dollars to me to confuse him.

How nice for you. I'm glad to see I'm helping you achieve your ambitions in life.

 

[SynapticBliss]

CaiNaM, funny link. "UR MOM" lol

-SynapticBliss

And, on the aside and very hush, hush, I think most of you people have eyes that are way better than mine. You see jaggies everywhere, places that I have to stare at, long and hard, to notice. In a perusal of a recent site I visited there were comparisons of NVIDIA and ATI top models with pictures with AAF this and that. I clicked each picture. Looked at the barel, the fire, the pant legs, the grate, the wall... A lot of tiny, tiny, tiny details people pan off as being incredibly important.

I don't care if I can't tell a leaf on a tree has a jaggy vein running through it where it could be a little more supple with a different card. Does the leaf stutter in the fake wind? Does it flow? Yes? Good. That's what I need.

Hey, look, the third hair on your avatar's beard, the one in the dimple on the chin, yeah, LOL, it's completely tweaked. You and your card are loosers! HAH!

No need to flame me. I won't notice, my card can't detect it.

-SynapticBliss

 

[ronnn]

beyond3d forums democoder is saying over 14000 on 3dmark 2003!

 

[DeeSlanger]

nu6800 14000 3DMk3 marks 12 pipecard 128 memory. oops

 

[SneakyStuff]

You know what I want to see tommorow, a benchmark using Quake 3 Arena, with 1024 x 768, and no AA/AF, how many FPS do you think a 6800u would fetch on that one

 

[vshah]

Originally posted by: CaiNaM
lmao....

check this out

too funny

i'm gonna die laughing......

Originally posted by: SneakyStuff
You know what I want to see tommorow, a benchmark using Quake 3 Arena, with 1024 x 768, and no AA/AF, how many FPS do you think a 6800u would fetch on that one

maybe it'd be cpu limited....
but thats why we still use q3 to bench...cause it scales so well.

-Vivan

 

[Pete]

Insomniak, B3D's forums may be an ATi haven ATM, but I don't think you can call their staff or their reviews biased. Otherwise, Dave wouldn't be benching an NV40 as we speak (25+ pages and counting, as nV probably wouldn't have given him a review sample).

I'll just say that the apparently system-limited nature of most of those leaked benches present a pretty strong case for 3DM03 or AM3 or any synthetic DX9 GPU-limited test, cheats or not, as the games Brent used apparently can't demonstrate the 6800U's potential. FarCry IQ is surprising and disappointing, though.

 

[Killrose]

Originally posted by: VIAN
"again, for all you people that think this is a fake"

The benchmarks that took is actually going against them in my opinion. Once they take benches with all the cards at the same settings, then you will cheer.

True, run'em exactly the same and the 6800Ultra is impressive. The COD benches are very impressive, and things can only get better with the next driver release. Anyone thinking about this card though should wait for R420 benches unless a nVidia based card is there only want.

 

[nRollo]

Originally posted by: Pete
Insomniak, B3D's forums may be an ATi haven ATM, but I don't think you can call their staff or their reviews biased. Otherwise, Dave wouldn't be benching an NV40 as we speak (25+ pages and counting, as nV probably wouldn't have given him a review sample).

I'll just say that the apparently system-limited nature of most of those leaked benches present a pretty strong case for 3DM03 or AM3 or any synthetic DX9 GPU-limited test, cheats or not, as the games Brent used apparently can't demonstrate the 6800U's potential. FarCry IQ is surprising and disappointing, though.

Hmmm. I don't know about that. Dave is a smart reviewer, and very knowledgable, but I think was at "3dfx Gamers" back in the day and never exactly a fan of nVidia. He's been picking apart their products for a lot of years.


If those HardOCP benches are real, they have a problem as well in that they should have shown all card at the the same settings, what they did leaves us speculating what the current fastest card would be at if the settings were equal. It's always been the battle cry of the ATI faithful that their darlings can run games at a little higher settings, now when Hard shows the reverse with a nVidia card, that doesn't matter at all strangely enough.

The Far Cry benches seemed OK to me, if you consider they're probably running at 32 bit precision, and for sure are at double the AA. Where's BFG to chime in the 2XAA on the 9800XT is a "butt ugly" setting?

Mostly, I think people are way off base thinking every time a new generation of cards comes out there's going to be some giant leap forward like the 9700Pro over 4600 was. I've been buying video cards to game with for a decade now, and this just doesn't happen very often. The last time it did was the Voodoo 1.
You just can't expect engineering marvels every year.