NV40 to have 222 million transistors

[NV55]

"NVIDIA NV40 to have 222 million transistors"
Posted on Saturday, April 10 2004 @ 18:33:52 CEST by LSDsmurf

Next week NVIDIA is going to launch its NV40. According to the Inq
this chip has 222 million transistors. The NV40 is produced by IBM on
a 130nm architecture. It will use the 60 series of drivers, currently
60.70.
As we said before Nvidia will push hard the fact that it is the only
firm that will have a Pixel Shader 3.0 model on the market and it will
focus its marketing efforts there. This means that apart from full
support for Shader Model 3.0 it will support model 3.0 Vertex Texture
Fetch/Long programs/Pixel Shader flow control Vertex Texture
Fetch/Long programs/Pixel Shader flow control and full speed fp32
shading. Let's hope everyone understands what this means.

There are plenty of other features of the chip which can do 16 pixels
per clock Color & Z or 32 pixels per clock Z-only, 64-bit FP Frame
Buffer Blending & Display, Lossless Color & Z-Compression and a new
Antialiasing approach called High Quality AA - Rotated Grid full MTR
(multi target rendering I guess), and accelerated shadow rendering.

Geforce 6800 Ultra has two power connectors and is, surprisingly a one
slot card for its reference card. It uses GDDR3 memory clocked at
550MHz but some partners might go even higher.

 

[Kentucky77]

And your point is?

 

[Andrew]

"NVIDIA NV40 to have 222 million transistors"

Ooh, transistor envy.

 

[Chocolate]

Did they say whether or not the cards would be overpriced pieces of shit
that create a 120 decibel swirling force-ten wind vortex just like the last
generation of nVidia cards did?

 

[rms]

Did they say whether or not the cards would be overpriced pieces of shit

that create a 120 decibel swirling force-ten wind vortex just like the last
generation of nVidia cards did?

And will it require a 480w powersupply, as the inquirer has also
reported? I suspect not, but it will require those two connectors to go to
separate rails.

rms

 

[Darkfalz]

Antialiasing approach called High Quality AA - Rotated Grid full MTR

(multi target rendering I guess), and accelerated shadow rendering.

Wow, only took them 4 years to catch up to the Voodoo5 in terms of AA
quality.

Of course, the Voodoo5 wasn't exactly playable in most games using 4xFSAA -
but it sure did look nice.

 

[Mr. Grinch]

Wow, only took them 4 years to catch up to the Voodoo5 in terms of AA
quality.

Of course, the Voodoo5 wasn't exactly playable in most games using
4xFSAA - but it sure did look nice.

I agree. I was also surprised that it's taken them this long to reach the
same level of FSAA that the V5 had way back when. Well, at least they've
finally done it, sort of. We still have to wait for the actual cards to
show up on shelves.

 

[Darkfalz]

I agree. I was also surprised that it's taken them this long to reach the
same level of FSAA that the V5 had way back when. Well, at least they've
finally done it, sort of. We still have to wait for the actual cards to
show up on shelves.

RGSS is the shit.

What method is ATI using?

 

[Tony DiMarzio]

I'm pretty sure ATI is still using OGSS

I agree RGSS/JGSS is superior AA tech.

 

[John Reynolds]

ATI's R3xx products have been using RGMS with gamma correction, far superior
to the competition's.

John

 

[Tony DiMarzio]

Thank you for that. Was unaware of R3xx's AA implementation type. However,
I'd argue that RGSS, while much more demanding of the GPU, still produces
better edge and texture quality than RGMS. Then again you can always combine
AF with AA on current hardware to achieve similar effects.

 

[Kentucky77]

I dont give a **** who has more transistors. I'm gonna buy the ****ing video
card that is the best. **** all you fan boys. I just dont give a ****. What
are you gonna bout dat.

 

[joe smith]

I dont give a **** who has more transistors. I'm gonna buy the ****ing
video

card that is the best. **** all you fan boys. I just dont give a ****. What
are you gonna bout dat.

How you know that the card you are going to buy is the best one of the lot?
On top of that, even if you by chance manage to buy the best card of the
lot, don't worry about it, it will not be in another six months or so.

I'm always buying the card that has the features I want to tinker around
with. Looks like nVidia is having the edge on that one, there is only rumor
out, but they say that the ATI won't be having VS/PS 3.0 on their next card,
just increased performance ( *yawn*, booooring... the performance been
alright for years for me already

Ofcourse your mileage may vary.. maybe you going to play Half-Life2 or maybe
DOOM III on the card of yours, who knows.. who cares.. but how you going to
decide which card is the best for use you are going to buy the card for,
when for example these two titles are still ways off from release when the
latest ATI and nVidia offerings hit the retail.. if you on the other hand
claim that you already have the software you going to be using with the
cards, then, RADEON 9800 (PRO/XT) or GeForce FX 59xx will do just great, you
don't need the latest cards from these vendors for existing applications -
they run fine already.

But your contribution to the world was very valuable, I learned a lot from
it, thank you for posting.

 

[joe smith]

Oh, by the way, depending on the definition of 'best', you might look into
ATI or nVidia workstation class cards like Quadro line from nVidia. The
prices are modestly cheap, few thousand bucks for a 3D card never ruined
anyone's balance. Then you will go into bed with your 3D card and make sweet
love with it (your bed is by the fireplace, right?).

 

[Chocolate]

How you know that the card you are going to buy is the best one of the lot?
On top of that, even if you by chance manage to buy the best card of the
lot, don't worry about it, it will not be in another six months or so.

I would have thought that following the application benchmark reviews would
give anyone a very accurate picture of any card's capabilities.

It's like the 9800Pro vs 59xx argument.

The 59xx was clearly shown to be an absolute waste of money. Of course,
there are still those who argue for nVidia until they're blue in the face,
despite the damning evidence.

 

[Kentucky77]

I still dont give a ****. You can talk, bitch think about stuff. And and I
still dont give a ****.

 

[Chocolate]

I still dont give a ****. You can talk, bitch think about stuff. And and I
still dont give a ****.

OK.

 

[chrisv]

(snip)

*plonk*

 

[patrickp]

I still dont give a ****. You can talk, bitch think about stuff. And and I
still dont give a ****.

So why do you bother reading this thread and posting to it, then?


patrickp

(e-mail address removed) - take five to email me

 

[joe smith]

I still dont give a ****. You can talk, bitch think about stuff. And and I

still dont give a ****.

You don't give a **** so much that you are going to **** yourself over when
you **** try to guess what the **** is the ****ing best gfx card. There is
no single answer, fanboy.