Re: EMC dominance
2-3 years is an eternity in today's technology marketplace. Curiously, that's also the lead EMC claims they have over their competition. Of course, as you note, EMC isn't exactly kicking back waiting for others to catch them.
This whole SAN vs. NAS "debate" is the sort of manufactured controversy that consultants and others with research to sell love. But it serves to distract the investing public from the real story, which is that server-attached storage is dead and this reality will inevitably result in the commoditization of the processor market. This is because the real customer value is in the data and once the data is independent of the processor then processing itself of necessity ceases to be the center of the enterprise.
Of course, full commoditization of processing requires two things: processor-independent data and processor-independent applications. While SAN and NAS are taking care of the first requirement, the second is being taken care of by the adoption of XML, Java, SOAP, and similar technologies which are enabling a new generation of processor-independent applications. Once neither applications nor data care about processor architecture it will be impossible to maintain anything resembling today's central processor margins.
Today, data is ahead of applications in achieving "escape velocity" from the processor-centric past. This is why it is still possible for the time being to ignore the implications of this trend for today's leading central processor vendors. Over the 2-3 year horizon of the ML report, however, processor-independent applications will begin to dominate their legacy counterparts. Ultimately, it is simply a question of relative growth rates. There is no need to throw out the old. Over time the old simply becomes a less and less significant part of the whole until it ceases to have any real relevance to business decision making. |
