"Don't know too much about how sales/marketing works, huh?"
Actually, I do. As you note, a better product only sells *itself* when it has a 10X advantage. Which, even the most optimistic evaluation of NGA doesn't place it.
So there are salesmen pushing the product to, say, HP. What HP wants to know is what justifies the price premium over competing AMD products. We will assume that NGA has a measurable lead over the top end AMD products at the time. Given the nearly exponential premium for high end, Intel just might be able to capture the high end from HP with no problem. But the vast middle is up for grabs. It all comes down to price/performance and what can Intel offer there? It takes time to build the "halo effect", so that isn't a factor until next year. Intel chipsets tend to cost more than equivalent AMD chipsets. Processors of equivalent performance tend to also. In addition, the bulk of Intel's production is going to be P4s until next year. They have the added disadvantage of requiring more power and, by extension, more expensive cooling. True, Intel can knock the 50 or so dollars off of their pricing, but cutting their ASPs by 30% means red ink. Because they can't price the high end stuff high enough to take up the slack. Unless the $3k plus notebook market takes off instead of the under $800, they can't make it up in notebooks. And servers, with no NGA for above dual sockets, there is limited opportunity there.
At this time, AMD doesn't have to push their product, although they should. It has already sold and is gaining share, both market and mind. Intel has to overcome that with product that won't exist for several quarters.
Market Data 
Markets 
