Hi Pete,
good to see y still around.
AMD has been doing development without a full fab for it for the last few years.
They muddled their way through even for decades. With mixed results. But that is history.
APM helps too. Yes it does. It helps to accelerate development, but it cannot substitute a development fab.
Having said this, AMD can pick up on using a fab for production and devopment at the same time again for the next node. It was definitely impossible while they migrated to 12''. Probably they need to do so - and will. So, I halfway agree. Because you still need to dedicate tool-time for development-wafer. Tools account for 80% of investment, shell, facilities and staff for 20%.
You could also argue that with sharing of process R&D costs with IBM and others, you don't need a full fab. Part of the initial development is at IBM, Fishkill and part at Fab 36. At no time was any full fab needed for development. Just part of a fab. That is AMD's way of R&D, efficient use of limited resources.
Umm. Are you aware if you say Fab 36 was partly a production fab you do accuse AMD for lying and subsidy fraud? In fact, the opposite is true: They had not enough heads, tools and waferstarts available to shrink its 90nm-architecture until now: Parts of the design of products they introduced last week are still 90nm. Let alone making its new architecture work.
Intel OTOH throws a full fab at it, running lots of wafers to try to get to usable yields fastest. They can throw gobs of money at the problems. Intel's way of R&D, quickest results at any cost. Problem is now Intel has less resources and problems cost more. Changing your R&D target from speed at any cost to efficient use of resources, is nearly impossible without big shake ups, usually at the top. Without a Max the Axe, inertia just rolls over any opposition.
I agree in what you say about resources: They skip Netburst and focus on one X-64 architecture. Obviously this frees resources.
Oh yes, and heads, with it. I've seen they trimmed headcount by a double-digit figure, as well.
K. |
