Re: We realize we have very different understandings of mobile-segment. The one i am talking about is 25W and 35W envelopes primarily, 65W-SKUs can only be utilized for the "desknote" subsegment of this market.
Again, I will challenge your notion of a "mobile optimized" process, by showing you 90nm parts that use the same 35W power envelope that AMD uses for 65nm parts.
By the way, AMD does not sell 25W mobile dual cores.
Re: The envelopes do not change from a performance-targeted to a power-targeted process - but the sweetspot of power-envelopes - corresponding with the volumes they have for these envelopes. Significant increase of mobile-volumes are visible for a while already and can be tracked back to process.
I haven't seen any evidence of a "significant increase" in 65nm mobile volumes, but even if there was, that would be a rather poor tradeoff for losing 600MHz worth of performance in the high end of the market, along with zero improvements in performance at equal power envelopes. Besides the dubious "volume" benefits (which I suppose are more driven by smaller die size), AMD's 65nm process is pretty much a bust in every other respect.
Re: Beyond, i don't think they are already below a sweet-spot of 45W which is still too high for mobile - i.e. the best of this is still to come.
I know the process will get better. But they have pretty much sunk a year into it (or it will have been a year before any new parts launch), and that's a long time to go with a crummy process transition.
News 
Market Data 
Discover 
