Did anyone find this release about chip growth weird:
semichips.org
I say weird because:
- There is no listing of this year's growth rates for all segments.
- In some cases there are cumulative totals; in others YOY.
- In some cases we get the industry sub-sector %ge - in most cases we do not.
- Most importantly - almost everything major is expected to do better than 12% yet the average is 12%? How does one pull off this trick?
----
Based on the SIA release, here's how I see 1999 growth:
Discrete: 7-8% (implication of the word steady)
Analog: single digits (implication of 'returning to double digits next year')
MOS Logic: 'faster' (how much? who can tell? Std cells 20% this year - implication of 'steady'?)
MOS Micro: average BUT later they mention DSPs are the industry's fastest growing line (and hence faster than 16% or even 19%????), microprocessors 16%, and microcontrollers 16%. Since MOS Micro is DSPs, MPUs, MCUs, and 'microperipherals' this means that for the entire category to be 'average' this year we need vodoo magic!
MOS Memory: 19%
Now how in the world can you get an average of 12% when most of the pieces are higher than the average? My best guess is that analog is a sizeable chunk of current revs and this is pulling things down. Still it must be huge and the growth rate pathetic if it is compensating for the 19% Memory growth, the 16% MPU growth, the fastest growing DSP segment, the faster than average MOS Logic sector. (Discretes we know are responsible for a 0.45% reduction in the combined growth rate.) Funny thing about this low analog growth rate is that the major analog companies (and the minor ones as well) are all doing well - as proof, note the backlog jumps and the sequential rev growth.
So a question to anyone who could help: Does anyone have the industry segment breakdown by %ge of revenues for 1998? Meaning, does anyone know what chunk of the total market analog was (30%, 20%?), and likewise for the other sub-sectors? |
