keith
Your pricing question inspired me to review my current model with emphasis on the ASP-forecast:
Although I still see AMDs portfolio trade up significantly, I currently do not see three digit ASP at any time for these reasons:
1. Global economic outlook looks muddy (to me). Growths in all developed regions more nominally than real, negative trade balances, increasing deficits, all that will likeley make a pessimistic environment for years to come. (Asia (except Japan) growing in the high single digits, positive trade balances, is a better place to be within the next years from an economic viewpoint.)
2. Consequently, I expect IT demand to grow slower than predicted. Unit-Shipments averaging up only in the single digits or very low double digits.
3. For PCs in particular, I see the fraction of users who seek more CPU-performance for their needs continue to decline. And I dont see any application for this year that would be able to change this perception except for some enthusiast niches. So, all the (unit)-growth I see is coming from mobile - better said laptop format. Not sure about serverspace. There is definitely a need for performance, guess Opterons will penetrate the market faster than every server processor ever did. But on the other side of this medal I see Xeons falling off a cliff as we speak.
4. Capacities at Intel are set for 15 percent market growth and current marketshares.
Conclusion: Intels cost advantages from 90nm/300mm will probably be used to lower prices (to spur demand, prevent from significant loss of market shares and to keep the fab utilized). Talking about 5% sequentially in every quarter for the next six quarters. For clarification, I am not talking about waterfalls, but net ASP declines.
Bottomline, I not only see no growth of CPU-market volume in Dollar-terms, but rather a likelyhood of a slight contraction.
5. Now, for AMDs ASP forecast, to trade up the portfolio in such an environment is still possible by means of server-CPUs, but probably not to the extent it would allow three-digit ASP as long as AMD delivers into all segments of X86. I am confident they will be in the 80s in the second half of this year again, maybe the 90s as well next year, but 100 Dollar will most probably be the ceiling: Actually, I see a crossover of Intels and AMDs ASP around 90 USD. Hard to say when, could be anytime between next year and 2007.
6. Just in case all that would sound too pessimistic: I sure hope it is. But even if it is not, there is still a lot of money in it for AMD: On a per share-basis, it is more than current estimates for Intel (which would appear optimistic anyway if the above is not far off what will happen.)
K.
Facts are irrelevant to the emotional brain
