ELSA hints GT206 and GT212

[Lukfi]

You gotta be kidding me. So, uh, the rumours about GT200 having crappy yields had a real basis it seems
Come to think about it, I haven't seen the original GT200 "naked" either. If this ~480 mm2 chip really is GT200, then does it mean they are adding NVIO into the official number?

 

[Tchock]

If GT200 was just 480mm^2 then it would be quite impressive.

But on the other hand, Fudo actually picked up a story that GT200 was also taped out in 55nm (no specific timeline) some time before the launch.

 

[Lukfi]

Yeah I remember that. So those are either 55nm GT200 chips, or dummies...

 

[Domell]

First question is if GT200 is really 480mm2, then why every source, every review and everyone said it`s 576mm2 (GPU-Z shows it too)? Is there any possibility that NVIDIA could lie with GT200 die size? If yes, the second question is WHY?

But the most important question is - Is there ANY review on the Net, which could confirm that GT200 IS 576mm2 and NOT 484mm2? Any review which shows GT200 GPU without heatspreader?

Here are some die size comparison. GT200 is 576mm2 here, even in ATI slide.

http://www.hardforum.com/showthread.php?t=1325165

 

[fellix]

There is a photo of G200's 300mm wafer floating around the Web -- count the integral chip pieces on it, then reverse the math expression found here and you will guess pretty well the actual die-size.

 

[Domell]

There is a photo of G200's 300mm wafer floating around the Web -- count the integral chip pieces on it, then reverse the math expression found here and you will guess pretty well the actual die-size.

Have you done that?

 

[Cookie Monster]

Anybody want to take off the IHS on their new shiney GTX280?

Btw, what is the die size of the NVIO chip on the GTX2x0 cards?

 

[fellix]

Btw, what is the die size of the NVIO chip on the GTX2x0 cards?

Pretty small for anyone to care about.

 

[Razor1]

http://images.anandtech.com/reviews/video/NVIDIA/GeForceGTX200/wafer.jpg

almost 12 across the full diameter

thats a bit more then 25x25mm

 

[Domell]

http://images.anandtech.com/reviews/video/NVIDIA/GeForceGTX200/wafer.jpg

almost 12 across the full diameter

thats a bit more then 25x25mm

So what the hell is that die? I don`t get it - then it seems taht it IS (or most likely is) GT206.

 

[fellix]

It's a 65nm G200 wafer -- 93~94 integral pieces out of it, and from that much less certified for the GTX280 SKU.

 

[Domell]

Yeah i know it`s GT200@65nm wafer but IF it really is, then what is that chip which is next to G80?

 

[Kowan]

Anybody want to take off the IHS on their new shiney GTX280?

Let me think about that...












NO!
ATM using my two F@H.

 

[Jawed]

Any chance the wafer of GT200s is an 80nm wafer?

Jawed

 

[fellix]

No Wai

 

[Razor1]

does 80nm come in 300mm wafers?

 

[Jawed]

does 80nm come in 300mm wafers?

What else?

Jawed

 

[Arty]

http://www.nordichardware.com/news,8065.html

 

[Jawed]

The picture from Tridam of the box of chips:

shows chips that have a ball layout that matches the chip pictures from DegustatoR:

http://forum.beyond3d.com/showpost.php?p=1223796&postcount=70

The Nordic Hardware pix:

http://www.nordichardware.com/news,8065.html

show a different set of balls. And the top of the chip shown by Nordic Hardware:

appears to be an exact match for the G80 core in this picture:

http://forum.beyond3d.com/showpost.php?p=1225339&postcount=77

as the "army of little components" surrounding the die is precisely the same in both cases.

So it seems there's a few lies going round.

Jawed

 

[fellix]

So, there is still no eyes-on over a genuine 65nm naked G200?