As an investment, Intel is a cash cow priced as a growth company, which is too high - they've have 80+ percent of a stagnant market, so the most they'll ever grow by increasing share is 20%. They've lost several fortunes trying to move into other industries, so their fate seems to be tied to the same old cash cow it's always been tied to.
Not sure I agree here... certainly, the market right now is stagnant, but once PC's grow again it will improve somewhat. In the meantime Intel has moved into communications and has taken over #1 share in that market (and that market has a large number of players with large growth potential). Problem is the communication market is also down... it'll come back, and Intel is poised to reap large gains. In the 64-bit space, Intel has the highest performing processor in the world in the Itanium 2 (which is manufactured on a completely obsolete 0.18 micron process). The 0.13 micron version of this chip will run circles around anything on release.
Meanwhile, their one competitor, AMD, has been gaining steadily in each of the past 1.5 to 2 year semi process cycles. It's not a straight line progression. Intel recovers some at the beginning of each cycle, such as last quarter, then AMD gains more ground.
I don't see a net gain here. AMD has been losing money for six quarters now, and will continue to bleed for at least two more quarters. AMD's 0.13 micron process is a dud with no headroom, and the need SOI merely to remain competitive, not gain an edge. The K7 is EOL, and the K8 is simply a minor revision of that core, with a couple of added pipeline stages, and an integrated memory controller that's going to be a validation nightmare.
Intel has also latched on to a rather nutty business plan, which involves abandoning it's X86 franchise to AMD, and attempting to set up new franchise with Itanium for the coming 64-bit CPU generation. If that backfires on them (and, so far, Itanium sales have been very discouraging) they could lose dominance in their primary business.
X86 will continue into the future for Intel... There's Prescott and Tejas on the horizon, both of which appear to be very strong x86 offerings. IA-64 is still not widely adopted, granted... but Merced was only a demonstrator, and McKinley is now beating the Power4 and all comers in performance... they're also cheaper... World class performance at a low price point is always attractive... it's why AMD made a bit of money while the K7 was still competitive performance-wise.
So, Intel's upside is quite limited, and the downside looks worse and worse as more news of X86-64 progress is released - with IBM's support from its premier database, DB2, being the most recent example.
What progress!?!?! I've seen a couple of programs, but no mainstream OS. AMD is trying to go with a "one size fits all" strategy. The problem is, "one size" usually "fits none." The K8 is going to be to weak to compete with the big iron, and it will have a lot of extensions and capabilities that no OS for the home desktop uses. Intel, however, will be turning on hyperthreading at or around 3GHz which will give the P4 a lot of added punch even using mainstream software.
AMD needs the K8 to survive, not "take over the world". AMD is in a very precarious situation at this point. They don't have a competitive product, and what product they do have is being produced on an uncompetitive manufacturing process. If the K8 and 0.13 micron SOI are not unqualified successes, then AMD will go the way of the do-do. |
