NVIDIA: Beyond G80...

Just disabled, still Conroe core. Allendale is the E4300, which doesn't have more than 2MB l2 cache.
 
Oh man Fudo really has no shame at all.

Fudo said:
We had a chance to test a Geforce 8800 GTX at its default 575 MHz with the old 97.92 drivers. We used Nvidia's beloved GPU DIP test.

In Cell shading the 97.92 does the job in 682,3311 msec and scores 403,3983 GFlop/sec. The same card, still clocked at 575 MHz for the core, but with the 158.16 drives now performs the same task in 584,9589 msec and scores 470,5480 GFlop/sec.

The third scenario included an overclocked G80 card clocked at 700 MHz, what can easily be the clock speed of G80 Ultra and the job was finished in 496,1952 msec and the card scored 554,7236 GFlop/sec.

This clown claimed he did the testing. Strange how he got the exact same scores down to 1/1000 th of a millisecond huh?

http://forum.beyond3d.com/showpost.php?p=969845&postcount=447 :LOL:
 
Sigh... it's not like it would have been less newsworthy if he just put: "We found the following on the Internet:..." beforehand.
 
Wow at least fudge the numbers a little bit if you're going to do that. Guess we know where he's been getting his information, as if we didn't already. lol
 
By "We" he means "The person who did the testing"... :LOL:
 
Trust me, if/when a $1000 overclocked G80 comes out I will be pointing and laughing right along with everyone else.

nV recently confirmed to some people that the GF8800Ultra with 768MB will be $999 at launch. They didn't say anything about clockspeeds.
 
i doubt many people will pay 1000 for a 650 to 700 mhz 8800. i also doubt that if r600 is all its cracked up to be that those clocks will give nvidia any speed titles.
 
nV recently confirmed to some people that the GF8800Ultra with 768MB will be $999 at launch. They didn't say anything about clockspeeds.

But what's the point? And it's going to suck for all those people trying to step up from a GTX - I doubt eVGA will have enough stock of this thing to cater to that crowd.
 
Another driver leaked, another large jump in the numbering scheme and gains in performance and... another codename:

Forceware 165.01
32bit XP

32bit Vista
64bit Vista

Supports all Geforce 6, 7 and 8 series products.



Newly listed product:

"G98" (not a typo :D)

Funny how nv is jumping in driver numbers without any real reason/benefit, whats next? forceware 289.18? ;)

What nv doing with driver numbers G98 can be a lowend G80 series gpu too :LOL:
 
Did you notice?

G84 features one TFA per TFU again... The basis for GF8900?

Yeah but wouldn't that border on the ridiculous in terms of texturing? That would be like a 45GT/s bilinear fillrate at 700Mhz ! I can't believe that G80's shaders are that texture bound right now.
 
Funny how nv is jumping in driver numbers without any real reason/benefit, whats next? forceware 289.18? ;)

What nv doing with driver numbers G98 can be a lowend G80 series gpu too :LOL:

Not exactly no benefit.
The new 165.01 driver is reported to be able to add some 25fps in CSS on average, for instance (compared to the 158.16).


And the GF6/7/8 official support, along with SLI on both DX9 and OpenGL in Vista is there (DX10 SLI would be pointless right now, as there are no titles yet).
Remember the 101.70 only had official support for GF8's, all the others were added via .inf hacks.
 
Not exactly no benefit.
The new driver is reported to be able to add some 25fps in CSS on average, for instance.


And the GF6/7/8 official support, along with SLI on both DX9 and OpenGL is there (DX10 SLI would be pointless right now, as there are no titles yet).

When AMD (i mean ATi) make some good ingame performance incrase with new drivers they not called it as Catalyst 14.8.
NV drivers vista performance still very poor, im not surprised NV is crying like a b**** for dx10 XP version ;)
 
Back
Top