POST OF THE DAY
Gorilla Game
Intel's Value Chain

Format for Printing

Format for printing

Request Reprints

Reuse/Reprint

By TinkerShaw
July 24, 2001

Posts selected for this feature rarely stand alone. They are usually a part of an ongoing thread, and are out of context when presented here. The material should be read in that light. How are these posts selected? Click here to find out and nominate a post yourself!

Pat,

Excellent way to articulate it. Intel controls the standard, Intel controls the market. If Intel sets a new standard for the market its value chain rallies around it because that is where the money is. If AMD tries to create new features its value chain puts it on the back burner to first service Intel's needs. What AMD wants only comes after what Intel wants. Certainly there are companies that are devoted completely towards AMD, but these are usually the companies that are not good enough to compete for Intel's business.

The P4 launch is a great example. Intel gets the DRAM industry to reluctantly, draggingly, but certainly, to invest in a new higher bandwidth memory solution. Intel also has the gall, yes THE GALL, to create the Pentium 4.

Why the gall? Because Pentium 4, despite its higher speed, does not perform any better, and perhaps a tad bit worse some would say, on legacy software applications. How dare they turn their back on their installed base of software! Evil Intel.

Well, this is the gorilla part. As Pentium 4 gains traction, new software will be written that is optimized for Pentium 4. Instead of text, graphics will be featured. Multitasking will become the norm, peer to peer will grow, etc. In two to three years Pentium IIIs and Athlons, chips made for the present installed base of software, will be left in the dust as relics. Everyone will need a Pentium 4-type architecture to efficiently run the new breed of software, as this software will become the norm and the installed base. The installed base will become as pass� as earlier versions of Lotus I-II-III.

Well, this is the monkey part (AMD monkey, not chimp, but Mike or someone can clarify the difference for you). AMD, after 5 years between architecture shifts by Intel, has gotten very, very, good at mimicking Intel's architecture. There is little difference between current software on an Athlon vs. a Pentium III, clock for clock. It doesn't matter much. AMD's suppliers, after 5 years, have also gotten very good, almost as good as Intel's suppliers, at producing quality, mass market, Pentium III based SDRAM systems. Thus, AMD, as is predicted by the manual, has caught up with Intel because Intel has taken too long to change the game again by changing its architecture.

Now monkey AMD (who has tried many times to move up to chimp but continually failed) is falling about a year (+ or - a few months) behind INTC, as AMD is that far behind INTC in moving to the .13 micron process. AMD had a fun time with the mature Athlon (Pentium III architecture) as INTC struggled to transition to its new architecture, but that time is over. Now AMD has to play catch-up again to INTC and it must do so at hyper-speed.

As the market situation looks going forward: INTC is aggressively pricing its P4s, INTC is also aggressively obsolescing the PIII and earlier P4s. INTC will be over 2 Ghz by Q3 of this year and moving at lightning speed. AMD will be nearly a Ghz behind in processor speed without anything, to my understanding until the middle of next year at the earliest. By this time, of course, INTC will be on a roll churning out 3 Ghz chips and AMD will be making its transition to a P4 type architecture, imitating INTC again (as it has already announced it would do with the SSEE2 instructions) and struggling to catch up in a nascent product for AMD and its value chain, yet now mature product for INTC and their value chain.

Now, there are only two ways this doesn't work. (1) The Pentium 4 totally fails and the world never moves beyond the current Athlon technology or basically the Pentium 3 architecture and software that goes with it. If so, computers and software become as commoditized as televisions; (2) AMD can some how start its own industry standard separate from INTC. It has valiantly tried to do this. Last year I did a report on this. Unfortunately, it again failed. AMD just could not get enough critical industry mass for this to occur.

So I hope that explains the process. It is a forward-looking game, a game of standards and value creation, and of being the entity that controls these standards at the time of this value creation pyramid.

Tinker

P.S. I have a bet on the Rambus board in that regard. That based upon the above theory, AMD marketshare will fall back to 16.5% by this time next year.