NVIDIA GT200 Rumours & Speculation Thread

Well, just like we're not talking about the free bag of gold I hid in Central Park, we tend not to discuss imaginary things, unless someone's quoting Fuad, of course.

 

should have checked before they started selling HD3xxx it was more then just one line

 

That's not the point of contention, which is actually semantical (the definition of "series") and this whole line of discussion is rather OT so I'll just stop now

 

Of course, I forgot to add the source, surprisingly is the same Nvidia and the information is public:

http://developer.nvidia.com/object/siggraph-2008.html

From it and for me, it seems pretty obviuos that GT200 contains a tesselation unit.

Please CJ post a slide to confirm this to those unbelievers (joking.. )

Other things:

At techarp they are talking about the same scores for Vantage that CJ, and that GTX 260 will launch or be available some weeks later, what do you think?

# One industry source told us that they tested their GTX 280 using 3DMark Vantage and it scored in the region of 4900 while the GeForce 9800 GTX scored only around 2300.
# NVIDIA will only launch the second model, the GTX 260, at a later date with some industry sources saying it should be around the middle of July. Apparently, the lack of competition on the high-end segment has given NVIDIA little incentive to speed up the release of the much cheaper GTX 260.

source:
http://www.techarp.com/showarticle.aspx?artno=550&pgno=0

For last, i forgot where i saw that EVGA will have a overclocked GTX 280 model with 675 Mhz for the core for about 599 euros.

 

From the link you gave:

It says nothing about a tessellation unit in the HW but instead mentions a technique that uses displacement mapping.

-FUDie

 

Where you see this:

I see this:

But I acknowledge this isn't the most clear string for extracting the info.
What about this from the same link:

and mostly this:

 

I wonder how much more than this:

http://developer.download.nvidia.com/presentations/2008/GDC/Inst_Tess_Compatible.pdf

the SIGGRAPH stuff is.

Jawed

 

Dunno if GT200 has a tessellation unit but all this new tessellation pipeline NVIDIA talks about is related to DX11.

 

That's what I was assuming.

Jawed

 

Already posted?
New slide from CJ, it seems.

 

Thanks. All that was already known though.

 

Can we please stop preemptive damage control? Thank you

 

A multi-GPU beating a single-GPU is kinda irrelevant, IMO.

 

Depends .

If the multi gpu set up is cheaper or similarly priced I don't see why its irrelevant.


there are pros and cons to both set ups I guess.

I personaly don't care , I will have a certian budget and whichever one is cheaper gets my dollar if performance is similar

 

What Damage Control? First of all, i am buying a GT280 !

Secondly, i hope you do not think i am serious.

Thirdly, that is what my now rare use of emoticons are for - to denote sarcasm
-- i.e "playing" Vantage

Do i need to really spell it out that i was making a bit of fun - at (1) AMD's *needing* X4 to almost-equal X2; and (2) my joke about the game included with the Futuremark Benchmark
-Now my fun - if there every was any - is completely gone. Perhaps i shall refrain from being sarcastic in the future. Spoil sport!

in other words - who cares about X2 beating X1 ? - in Vantage; a synthetic bench that AMD recently does well in

 

It's okay - It's not the first time the limmited bandwidth of print has failed to properly represent a person. Print can only be so verbose.

 

UGH...

 

He's referring to the fact that you are already trying to justify a purchase (that you haven't even made yet!) because all the evidence points to another card being faster than the one you plan to purchase.

Everyone? I don't care (and I would imagine most will agree with me) if it takes AMD 3824792 cores on one PCB to beat Nvidia (as long as it's not stupid expensive).

You don't. Everyone got it but no one thought it was funny or even believed for a second that it was a good point.

 

I dont know about you guys, but id prefer single GPU solutions over multi GPU solutions. There are just too many headaches involved with the multi GPU solutions (+ its limitations) even if it performs couple more framerates than a single GPU solution.