Wait, the same Charlie that says TSMC has 40nm yield issue but it is not AMD's fault. But when Charlie says TSMC has 40nm yield issue, he says nvidia is screwed. Any moron can read Charlie's articles and tell he is Bias. So that is what journalism is about. Or is that just marketing.
A process node yield issue is rarely related to the quality of the process, rather the architecture that is being shrunk.
It is commonly believed -- as it has been reported by too many sources far too many times to be 100% wrong -- that NVIDIA's current architectures have exceptionally high leakage past 55nm.
There's plenty of evidence to support this fact: The G212, a 55-to-40nm die shrink of the G200b, is dead and gone. It's never coming back. The G214 is so dead that it doesn't even exist in ForceWare INF files any more; the G215 took its place and it's 8 months late.
The remaining scheduled 40nm parts? Eight months late, and no closer to release (G216 & G218). The 216 and 218 were supposed to be here now. Not only are they not here, that hasn't even been word that they <i>might</i> be coming on the planned schedule.
Even during NVIDIA's big spring analyst call, all talk was centered on the GT300, and nary a peep was mentioned about the firm's 40nm ambitions. Hey, what do you do if you want to impress analysts? You tell them that your 40nm GPU designs rock and are on track for their scheduled release dates. What do you do if you don't want your stock to get blitzkrieged? You don't mention anything about your failing process.
At least we got a consolation prize: The GeForce GTX 275 on the 55nm G200b core instead.
So now we come to the GT300, the first player in what was actually supposed to be the grand entrance of a process node NVIDIA perfected on lesser GPUs. It's late. It's probably delayed. If it makes it out in 2009, it'll be a $650 board with so limited a run that reviewers may be the only ones who ever get to touch it.
There's no bias here. NVIDIA has left a raft of evidence to shore up the discussions. You don't ditch an entire quarter of your roadmap, kill two GPUs, and delay 3 more if everything is fine.
TSMC is not the issue. It's NVIDIA. I guarantee that ATI will ship in volume.
BuddyJ
Dept. of PropagandaOKC Icrontian
Comments
It is commonly believed -- as it has been reported by too many sources far too many times to be 100% wrong -- that NVIDIA's current architectures have exceptionally high leakage past 55nm.
There's plenty of evidence to support this fact: The G212, a 55-to-40nm die shrink of the G200b, is dead and gone. It's never coming back. The G214 is so dead that it doesn't even exist in ForceWare INF files any more; the G215 took its place and it's 8 months late.
The remaining scheduled 40nm parts? Eight months late, and no closer to release (G216 & G218). The 216 and 218 were supposed to be here now. Not only are they not here, that hasn't even been word that they <i>might</i> be coming on the planned schedule.
Even during NVIDIA's big spring analyst call, all talk was centered on the GT300, and nary a peep was mentioned about the firm's 40nm ambitions. Hey, what do you do if you want to impress analysts? You tell them that your 40nm GPU designs rock and are on track for their scheduled release dates. What do you do if you don't want your stock to get blitzkrieged? You don't mention anything about your failing process.
At least we got a consolation prize: The GeForce GTX 275 on the 55nm G200b core instead.
So now we come to the GT300, the first player in what was actually supposed to be the grand entrance of a process node NVIDIA perfected on lesser GPUs. It's late. It's probably delayed. If it makes it out in 2009, it'll be a $650 board with so limited a run that reviewers may be the only ones who ever get to touch it.
There's no bias here. NVIDIA has left a raft of evidence to shore up the discussions. You don't ditch an entire quarter of your roadmap, kill two GPUs, and delay 3 more if everything is fine.
TSMC is not the issue. It's NVIDIA. I guarantee that ATI will ship in volume.
<object width="425" height="344"><param name="movie" value="http://www.youtube.com/v/4WxDesCYVmM&hl=en&fs=1&"></param><param name="allowFullScreen" value="true"></param><param name="allowscriptaccess" value="always"></param><embed src="http://www.youtube.com/v/4WxDesCYVmM&hl=en&fs=1&" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="425" height="344"></embed></object>