Nvidia GT300 core: Speculation

Status
Not open for further replies.
What makes you so sure about it, neliz? Any data yet available for GT215 and the like?

Well, you heard Degustator about the GT215 performance, that's basically in line with other rumored specs.

My yes statement has just been shot out of the sky though, suddenly that "December review" thing by TGDaily makes a lot more sense, no bonuses for the GT21x team.
 
My yes statement has just been shot out of the sky though, suddenly that "December review" thing by TGDaily makes a lot more sense, no bonuses for the GT21x team.
Could you elaborate a bit? I feel like I missed something... (Sorry, waking up late and groggy)
 
http://digitimes.com/news/a20090918PD210.html

Huang is expected to hold talks with TSMC chairman Morris Chang to discuss contract prices for 40nm production, according to sources at graphics card makers. He is also expected to check on TSMC's 40nm process, for which TSMC has claimed yield rates have already risen from 30%, to 60% in July, and defect density will drop to only 0.2 in October. Huang will also check the progress of its upcoming GT300 GPU which is expected to launch in December.
 
for which TSMC has claimed yield rates have already risen from 30%, to 60% in July,

Looks good at first glance but is that across the board or just for ATi design?
I assume yields are affected not only by the size of die & complexity but also by differences in design?
 
I don't know what kind of reasoning TSMC used when claiming such figures, but I have the feeling that it would had been better if they had claimed something along the line that yields have improved by up to 50% depending on case. Now if one chip had 20% it might be 40% today and naturally another one at former 30% being at 60% today.
 
Because of the exponential relationship they can't give a flat percentage for that either Ailuros, still depends on design.

Does anyone know the average percentage of non redundant circuitry on a modern GPU so I can get some idea what 0.2 defect density will mean in October?

PS. that's per mm2 I assume? That seems awfully fucking high.
 
Just saying it would not work like this ... "20% it might be 40% today and naturally another one at former 30% being at 60%". You do not get a flat percentage increase in the number of working dies per wafer. It will probably be closer to, if 30 goes to 60 then 10 goes to 40 for yields.
 
Just saying it would not work like this ... "20% it might be 40% today and naturally another one at former 30% being at 60%". You do not get a flat percentage increase in the number of working dies per wafer. It will probably be closer to, if 30 goes to 60 then 10 goes to 40 for yields.

I understand what you are saying; but an "up to" X figure claim includes all possible case scenarios and you mention naturally the highest possible increase because otherwise you wouldn't make a marketing claim LOL :LOL: Just replace my initial 50% with the highest achieved percentage and mark it as "up to".
 
Well the highest percentage increase in yields will undoubtedly come from the G300 (not to say I lend any credence to the 1.7 rumour, just saying it's probably the chip with the largest amount of non redundant circuitry they are running through the process at the moment).
 
Well since Jensen is visiting TSMC they might as well abuse that Hitler movie video again, where the officers will be the TSMC folks :devilish:
 
http://murobbs.plaza.fi/1703676440-post252.html

Translation:
My NVIDIA contact had this as his Messenger personal message: "That's it? Time to wait for NVIDIA", so I inquired him for how long we have to wait.

The replies:

"Coming soon"

"Put it this way, you'll have plenty of time to write your christmas list"

I told him I was wrapping up my 5870 article:

"hope you're going to say it is worth waiting for what NVIDIA have coming up"

I told him that I have not heard anything from him or from NVIDIA, so it's a bit hard to tell the readers what to wait for.

"I can't give you any information, you know that, although I am sure it will not be long to wait now"
 
Status
Not open for further replies.
Back
Top