Computer designs are all about platforms, and CPU's by themselves are only part of the equation. When it comes to chipsets, AMD has been largely reliant on 3rd party vendors to get its platforms into the marketplace. Nvidia, VIA, SIS and Uli have all wrangled back and forth over the years trying to develop the most cost effective and performance oriented chipset designs for AMD. AMD's own foray's into the world of chipsets have been out of necessity, and quite brief. Keep in mind that this is truly a David vs. Goliath situation, and always has been. AMD's experienced its own ups and downs over the years as a result of not having total sway over its own marketability, but over the years as their products have improved, so has its 3rd party vendor support. Today we have a mature and stable "platform" for AMD, and it's largely because AMD doesn't (and truly can't) dominate every aspect of their product platform. Of course it helps that they have a solidly engineered product too, but if they didn't we wouldn't be having this conversation either would we?

Intel today is in every sense of the word, a platform vendor. Intel has a vested interest not only in its CPU's, but in its lucrative chipset business, and let's not forget Intel's past attempts at tying up the DRAM pie (Rambus anyone?) or at altering form factors (BTX). This platform approach has been with Intel for quite a while, though its first important incarnation for my purposes, was the P4/Rambus period. It's not always been a successful approach for them either. Being the 800lb gorilla in PC universe doesn't always mean you're successful, or even right. Shortly before this period and up through today, Intel has focused on trying to get consumers (whether they be mom and dad, or corporate IT wonks) to perceive Intel as the superior brand. There are some who would argue that this coupled with a tinge of arrogance and blindness, has led to some of the stumbles we've seen Intel take over the last eight years. And some would say, the only reason why Intel has survived it has been because, well this is Intel we're talking about. Massive companies, and the massive industries and infrastructures surrounding them also have massive amounts of intertia. Why take the chance on someone else, when sticking with what works has always worked for you.

Don't misunderstand me. I'm really not trying to paint AMD as the good guy or Intel as the bad guy. It's "fanboy'ism" to paint a complex situation in such a black and white manner. Contrary to what some may focus on in this business, AMD and Intel are corporations. Interested in selling product. Interested in expanding their scope. Interested in expanding their markets. Interested in showing good margins to their shareholders. When it comes right down to it, this is all about money. Still good engineering definitely is the order of the day towards making all of these things happen to their fullest potential. Good decision making is also worthwhile.

I'm an idealist at heart. Not only that, I truly feel that a competitive playing field benefits everyone involved, *including* the competitors. The fact that Intel has been metaphorically "contemplating its navel" in the past is pretty obvious, at least to my way of thinking. Has their about-face been because they've learned from their past mistakes? And did they learn enough? The *bigger picture* here, at least right now at the actual launch of Core 2 on the desktop, gives me pause for at least a little bit of concern. Some might take exception to my thinking here, and that's fine. Devil's advocacy is something I enjoy doing because forcing oneself to be a little skeptical sometimes helps you stop long enough to see the bigger picture. Some of my leanings here are conjecture, but I think they are worthwhile issues to ponder, so bear with me.

Core 2's desktop chipset of choice is the P965 series. Sure, later iterations of i975 will do too, but Intel is banking on P965. As things stand, there are indications that at least in the short term there may be supply issues with the chipset for a couple of reasons. The most interesting to come of light, seems to be the southbridge's initial lack of IDE support. Some may scoff, or say "who cares, IDE is dead, big deal." Tell that to system integrators and board manufacturers who would have to eat the cost of expensive SATA ROM drives, or the additional expense of a few more $$$ for a 3rd party IDE controller. That few $$$ is a *huge deal* to mainboard vendors, who's profit margins hover in the low single digit percent range. It's also a huge deal to builders and system integrators, with the cost difference between and IDE and SATA ROM drive being quite HUGE. Sure, there's blame to go around here. ROM drive manufacturers have stubbornly avoided SATA, even when the cost of implementing it can't be tremendous. Even Plextor, the only vendor I'm aware of actively selling SATA drives, simply welded on a SATA to IDE bridge to their drives. What gives? Even if it's not entirely Intel's fault, it's also been Intel's decision to drive and dominate its own chipset market. Did they not see this coming? Hopefully the retooling of the southbridge (apparently Marvell Semiconductor is doing this for Intel) hasn't been as drama packed as the news that's been trickling out indicates.