Jim, Re: X-box chips @ 0.18 / 0.13
I think that Intel's 0.18 processes should be more than capable of providing the volumes needed for the X-box. 10M/year? Isn't that just a week or two for a mega-fab? :^)
WRT Intel's 0.13 production capacity, it seems like this is being committed to all fronts: notebook, low-power server, P4, and desktop Tualatins (?). This seems like a pretty wide front for a new and only gradually expanding process. I don't think the X-Box will need 0.13 micron chips anytime soon unless MS plans an aggressive product cycle, perhaps moving everyone to a Win-XP derivative? <g>
Intel (and AMD) will both face the question of what to do with their fabs which work fine at 0.18 as the CPU market moves to smaller geometries. To a certain extent, these can just be shifted to other products (chipsets, flash, etc) rather than going through the cost of taking portions out of service for refitting. Also, if these fabs have limited capabilities in other areas (such as quality of clean room air) then it might not make economic sense to re-equip. Going forward, the 0.18 equipment is "free" as I would doubt there would be much resell value.
I thought most of Intel's $7.5B in capital investments this year was for new fabs rather than a significant portion for equipment upgrades in existing fabs.
I think AMD needs to move forward with Fab 35 as soon as possible as well as preparations for an expansion of Fab 30. Well, I guess Fab 30 still has some empty floor space but it does take some time to design, permit, and build even in a regulatory supportive environment (as I believe Dresden to be).
-PT |
