The first question you have to ask about Cisco’s unified computing system (UCS) is, why would Cisco enter the server market now? After all the last decade has seen far more companies exit this market than enter it? Surely it’s so commoditized that’s there little left to do but find ways to make a server cheaper than everybody else!
Cisco would dispute both of these arguments. First they believe that they are not entering the server market they are creating a new market, that of unified computing systems. Second, they argue that unified computing, which emphasizes the integration of the entire infrastructure combined with ubiquitous server virtualization, offers plenty of scope for innovation at all levels, and the high-margins that a differentiated product brings.
I can certainly see the strength of the latter argument; the combination of server virtualization, server platforms optimized for server virtualization workloads, sophisticated I/O virtualization technologies and unified network fabrics open up tremendous opportunities to optimize the compute infrastructure.
What I’m not so sure about is the first argument, i.e. that this is a new market, not just part of the natural evolution of the commodity server market. My feeling is that existing vendors are already heading down the same path, particularly with their blade servers (HP would argue that they are already there). So, my feeling is that Cisco may have got the jump on some of the vendors for a few months, maybe even a year, but the other vendors will respond.
Of course, the real motivation behind this is that Cisco needs a new market to expand in. The halcyon days of explosive growth in enterprise networking revenues are probably gone, and grabbing a slice of x86 pie probably looks pretty attractive right now, lets just hope that Cisco’s eyes aren’t bigger than their appetite!
Posted by: Nik Simpson


Why now? Two reasons from my perspective:
(1) As you point out, it's a natural adjacent growth market for them, as Chambers said in his opening remarks.
(2) Also, per Cisco's repeated remarks, this is a *differentiated* offering. It's not about selling "blades". Blade vendors who think that's the point are destined to be forced to exit the market. This point of differentiation is that it's about how holistic data centers are managed. And Cisco thinks it can own that space.
Posted by: Ken Oestreich | March 16, 2009 at 04:09 PM
NiK:
We see a very distinct market transition around virtualization and we see a couple of unmet needs. First, we see a need infrastructure that is designed with virtualization in mind. Remember, the reason we have server virtualization in the first pale is because of the inefficiencies of traditional silo-ed infrastructure and legacy server design. Secondly, and perhaps more importantly, we need a better way to manage all these virtual resources and their inherently dynamic nature. So, the "why now" is because we see right now as the inflection point to go to market with this solution.
Omar Sultan
Cisco
Posted by: Omar Sultan | March 16, 2009 at 05:16 PM
Hi Nik, how you doin?
The timing is interesting. The UCS vision is a long ways from being implemented and it's going to take years of work to get there. My guess is that the answer to "why now?" is that this was the earliest Cisco could pull everything together to make a credible story other than "our core business is starting to suck".
Posted by: marc farley | March 16, 2009 at 10:13 PM