Re: The real issue is what happens *next* year when Intel
opens the 65 nm floodgates and AMD starts ramping
its new fab. I think AMD will be able to withstand this
early on better than in the past because of its good,
competitive product line in the server and desktop PC
segments (although this might be a pyrrhic victory if
the mobile sector keeps growing quickly at desktop's
expense). But once Conroe starts shipping it could
be back to the bad old K5/K6 days for AMD.
I think you'll start seeing this happen before Conroe launches. It will be like the Willamette to Northwood transition, where Intel successfully ramped Willamette in spite of it being an inferior product to Athlon XP, but when Intel moved to Northwood, they cranked them out faster than AMD knew what to do; they quickly became uncompetitive and ASPs plummeted. By mid-06, Intel will be cranking out large volumes of Yonah, Presler, Dempsey, and Cedar Mill at prices competitive enough to ramp the fabs at full capacity. The wheel will already be in motion before Merom, Conroe, and Woodcrest even hit the market. I think the effect on the latter will be to help Intel restore margins that they had to sacrifice with the former in order to ramp volumes. So basically, H1 will be sacrificing margins to ramp volumes, and H2 will get the best of both worlds.
Market Data 
Markets 
AI
