Joe, Re: By the time Mentocito is released at 90nm, Intel will be readying 65nm dual Yonah.
And by the time Intel can qualify a 65nm dual Yonah for high end server related tasks, and add the appropriate CPU level RAS, there will probably be a 65nm Montvale.
The point is that your comparison is apples to oranges. If the market was ready for a TTM IPF part without any server related RAS, Intel could deliver such a product much faster.
You have to realize that we are comparing two different worlds. You can't say that Dothan would be a better server design because it does better in SPECint, or because it will reach 65nm before IPF. The reasons behind those achievements is exactly what prevents it from being a successful design in high end servers.
Market Data 
Markets 
AI
