Technocrat. Thanks for the time to tell us of your FC research.
Any thoughts on this post from the EMC thread?
Message 10454283
EMC, HWP, and the future of the datacenter
The unfolding of events between EMC and HWP should come as no surprise to EMC shareholders and in fact is further evidence of EMC's emerging gorilladom.
As I've stated before, what we are witnessing is a polar shift in the "center of gravity" within the datacenter. Historically the processor (in the guise of the Almighty Mainframe) was the centerpiece of the datacenter budget and storage was something that was, literally, a "peripheral". Today corporations are increasingly recognizing that their data is at the center of the enterprise and the care and feeding of that data is the basis for all datacenter budgeting and resource planning.
After 2005 as the IBM mainframe era enters its twilight we will see a complete reversal of the traditional hierarchy in which storage will become completely ascendant and the mainframe will devolve into massively parallel application servers which exist solely to perform manipulations of the data. The architecture of this 21st century datacenter will consist of a central Storage Area Network (SAN) woven together with optical fiber and connected to the intranet/extranet/internet and attended to by a halo of clustered processing nodes consisting of between dozens and thousands of 64-bit enterprise-class microprocessors. Management will be completely data-centric and network-centric and processor vendors will compete for attachment rights to this SAN.
The reason the enterprise market has been so highly prized by all computing vendors is because of the lush margins it commands. In an era of around-the-clock e-business where a single outage can ruin a company's financial performance for a quarter only the foolhardy gamble with anything less than top-tier suppliers. But this shift has two implications which every investor must keep in mind when considering the long-term prospects of the various competitors.
First, the highest margins flow to the center. Historically these existed in the processor market but they will migrate to the SAN. The implication for both EMC as well as for traditional processor vendors should be obvious and it explains why IBM, SUNW, HWP, etc. are at pains to establish an independent storage presence for themselves. But as I previously observed, EMC's leadership in this increasingly-important (and profitable) segment is unlikely to be challenged by competitors who are dragging a processor bias along with their new found storage mission.
Second, the enterprise processor business will become commoditized. Any company whose business depends on selling high-margin iron is in for a painful transition as the premiums they formerly enjoyed are shifted from the processor to the SAN and they are forced to compete as the new "peripherals" to the SAN. This, of course, is the opening that Wintel has been waiting for and is the reason why no one can afford to dismiss this threat to the current world-order in the datacenter. While the residual margins available to processor vendors in the SAN world will look meager by the traditional measures of the IBM mainframe world, they will look lush indeed compared to the wafer-thin margins of the desktop and departmental-server world that the new competitors will come from.
All of which explains why EMC should be preserved as a core technology investment for many years to come.
If this post is correct what does it mean for FC switches?
Any thoughts out there on this post? Thanks in advance. |
