What you are overlooking is Intel's ambitious plans for the P4. Intel may be bringing capacity online, but at least in H1 2002, this will be offset by growth in average die size.
---What you seem to be overlooking, is that the capacity being brought on-line, is at 0.13u, not 0.18u. The smaller design size that will be manufactured at Fab22, plus the continued ramp of Fab20, will more than take care of any capacity issues arising from larger die sizes.
By contrast, AMD will probably finish ramping wafer starts by the end of 2001 with 20% of this ramp being 193nm equipment in Q4. This capacity will only begin to ship at the end of 2001, with perhaps 17% of Fab30's capacity not being realized in shipments until Q1 2002.
---It has been widely reported that there is a delay in the shipping of 193nm Litho Steppers. I doubt that AMD is at the top of the waiting list. Your presumption that "20% of this ramp being on 193nm equipment in Q4" is wishfull thinking at best.
Additionally, AMD will not be moving to a large core than the Palomino until Sledgehammer ships in late 2002. Until that time, AMD will be experiencing perhaps as much as a 50% increase in unit volume from Fab30 in 2002 just from the migration to 0.13-micron.
---It's interesting how at the beginning of your message, you ignore the volume increase from intel's transition to 0.13u, yet here you tout it for AMD.
Your statement is assuming of course, that AMD can get their 0.13u Process working during that timeframe. Not a quarantee at all. In contrast, intel has been manufacturing 0.13u in Oregon's Fab20 since the beginning of this year, and by the end of this year, they will be starting the shipping of this same technology in their 2.3 Billion Dollar Fab22.
Your predictions seem to be based on "If", mine are based on "What Is". I'll go with mine.
Semi |
