Capacity revisited yet again, and again.
The reason I keep harping on this issue, more than anything else, is that I consider it to be the most important thing effecting AMD's health right now.
It occurred to me that in the last post I hadn't speculated on how AMD might be handling the capacity constraint issues. More to the point, how is it that we haven't seen more evidence of capacity problems in the general marketplace?
The overall answer is probably superior management. In specific the fab30 plant manager and the head of marketing along with their staffs have evidently been doing a superlative job of matching capacity with demand.
As to how they may have been managing to a scarce capacity situation; a couple of things seem to continually prop up as complaints on this board and others. I'm thinking of the seeming unwillingness of the OEMs to in-fill and expand product lines using AMD processors and the lack of advertising of AMD products.
Perhaps the lack of acceptance by the OEMs has had more to do with AMD not being able to supply additional processors than a lack of interest by the OEMs? Also, not advertising keeps a lid on demand, while reducing expenses. These things come to mind as possible ways of managing the supply issue, but I'm sure there are other ways being used that I'm not aware of such as restricting promotional incentives etc.
For the most part demand seems to be an OEM issue rather than an ultimate user issue. How many, even technical people, really understand the advantages of say an on board memory controller? Or more importantly, given that the processor is just one part of a system, how can the average consumer easily compare the advantages of one processor to another? Given these problems it seems most likely that the ultimate consumer buys a particular PC based on other factors than the processor, primarily relying on their general feelings about the OEM/seller. Having said this, it is notable that there is a small, but growing, body of PC purchasers that do understand some of the processor issues and choose accordingly.
So from AMD's point of view their primary customer is really the OEMs and these guys are probably concerned with issues that effect cost and supply more than anything else. What it logically comes down to is if AMD can't guarantee supply the OEMs won't go through the expense of expanding product lines. This is probably the primary means AMD has used to restrict demand, just coordinating effectively with the OEMs. End consumer demand by the technically aware just isn't that big a factor compared to OEM demand. To me that's the reason we're not seeing processor shortages in the market, along with AMD's successes at managing other than the OEM demand.
06 should be interesting as it looks like processor supply from AMD and INTC will exceed growth in demand by a large margin. I expect INTC to be the big looser in this, as forces already in place will continue to expand the demand for AMD processors. Indeed, given the growth dynamics AMD has exhibited over the last couple of years it seems most likely that growth will accelerate as projects that were put on hold by the OEMs due to supply constraints are now allowed to go forward. End demand should also be effected as AMD name recognition continues to improve through increased advertising and business becomes ever more aware of the AMD advantages. One other thing that could effect AMD greatly is the law suit. As this thing winds through the courts the media should have a field day. Good lawyers are basically showmen, and this suit if handled properly could be a thing JP Barnum would have embraced.
What I expect to happen is more of what we've seen over the last couple of years. Namely, AMD will continue to take a disproportionate share of the increase in new demand coming from expansion, largely in the markets of China, India etc. Secondly the OEMs will continue to expand and develop new product lines using AMD processors. Third sales of servers and laptops will lead AMD growth percentage gains and due to their profitability effect revenue even more so than currently.
Big unknowns include the effect the PIC will have on sales and miscellaneous expenses. Overall though, embedded processor sales of Geode and Alchemy products will probably continue to increase and even though Spansion will have less of an impact it should still produce a positive rather than a negative effect in 06.
AI
