[rescue] Apple to ditch IBM, switch to Intel chips

[Barry Keeney]

On Sun, 19 Jun 2005, Patrick Giagnocavo wrote:

> On Sun, Jun 19, 2005 at 08:00:24AM -0400, Lionel Peterson wrote:
> > From: "Patrick Giagnocavo" <patrick at mail.zill.net>
> > >
> > > But wouldn't it be logical that the price would drop as volume
> > > increased?  Surely they would be able to afford to bring up another
> > > fab, or run the fabs they have at 110% for a while, if it meant
> > > selling 3 million or more chips to Apple?
> PatrickG ^^^^^^^^
>  
> > How do you run a 24x7x365 fab at 110%? On a leap year you get an extra what, 
> > .25% from the extra day (29 Feb)...
> Lionel ^^^^^^^
> 
> Overtime/extra shifts.  Yes it seems like a funny way to represent it,
> but that is how semiconductor manufacturers represent it, as a %-age
> greater than 100.

  I've got this info from friends and family who work at the
local Intel fab's.....

  Doesn't work that way. With a fab that runs 24x7, 350+ days a year, 
every station has someone working it, the equipment is already running
as fast as possible. 

  More chips they make, the more profit they make. If the chip is in
demand they'll produce them as fast as possible. If the demand drops
they slow the production rate in order to keep the fab running 24x7.
If demand drops to low, they'll either move it to a smaller fab or
maybe produce the chip for only part of the year then switch to 
a different chip the rest of the time.

  Only way to increase production is to build/convert more fabs to making
the chip or not to shutdown on the few days a year they don't work.
(the week between Xmas and new year, maybe thanksgiving or July 4th) 
It can take between 24 and 48 hours to get a fab up to 100% after a
shutdown.

Barry Keeney
Chaos Consulting
email barryk at chaoscon.com

"Rap is Square Dancing gone terribly, terribly Wrong...."