Michael Capellas, the former CEO at Compaq and since May of this year the CEO heading up the Acadia cloudy infrastructure partnership between Cisco Systems, EMC, and VMware, has been making the rounds, talking up the partnership and its Vblock virtualized server and storage stacks.
In a column over at CNET, John Webster, a senior partner at IT analyst firm Evaluator Group, has revealed some information that Capellas shared with IT analysts about the progress of the Acadia partnership through early December. Server wannabe Cisco and enterprise storage behemoth EMC are the two major forces behind Acadia, with EMC's VMware virtualization subsidiary and chip maker Intel also kicking in money and technology for the Vblock stacks.
The Acadia effort, now known as the Virtual Computing Environment (VCE) coalition, was announced in early November 2009, and basically takes Cisco's "California" Unified Computing System, with its Xeon blade servers and converged Nexus switches, and adds in EMC external storage, VMware for server and network virtualization, and BMC Software's BladeLogic system and application management software, which augments the network and system management tools that Cisco has embedded into the UCS top-of-rack switch. El Reg drilled down into the Acadia partnership and the initial Vblock configurations here.
Capellas has told analysts that the VCE joint venture has 65 customers and that the average Vblock deal is $2.5m. The coalition is seeing a month-to-month growth rate of 40 per cent and has a pipeline of prospective clients and proof-of-concept buyers who could upgrade that stood at $1.2bn as of the beginning of this month.
If you do the math - average deal size multiplied by customer count - you're talking about $162.5m in revenues. And if you realize the base Vblock 0 setup has a few blades and all the storage and switches and costs hundreds of thousands of dollars, the Vblock 1 setup has 16 to 32 blades and costs from $1m to $2.8m, and the Vblock 3 setup has 32 to 64 blades and a price that starts at $6m, then you also realize that it doesn't take very many blades at somewhere between $62,500 and $93,750 at list price to hit that $162.5m.
Assuming an average configuration - the midpoint of the Vblock 1 setup - and a 30 per cent discount, you're talking $55,000 per blade on average for a Vblock setup. That's around 3,000 blades across those 65 customers, perhaps supporting something on the order of 280,000 virtual machines in total.
While those numbers may sound impressive for a "startup," as if you could call something with these four (well, more like 2.75, really, including Intel which has only a modest role) a startup, there is no way that the 40 per cent month-to-month growth rate can hold up in 2011.
There is every reason to suspect, in fact, that month-on-month growth for the VCE partnership has been slowing dramatically through 2010 from some very high, triple digit numbers earlier in the year. This is a perfectly natural ramp for a new and innovative product. If that 40 per cent rate held, by this time next year, the VCE partnership would have $9.2bn in sales and, provided the potential customer base for Vblocks was growing as fast as sales, a $68bn order pipeline. This just ain't gonna happen, obviously.
It is hard to say how fast VCE revenues will grow in 2011, but something closer to 20 per cent per month in early 2011 and then cooling down to maybe 10 per cent by the end of the year is a reasonable guess. Should that happen, VCE could break $1.2bn in sales in 2011 and clear its pipeline, and at the same growth rates, it would still have nearly $9bn in pipeline built up.
The VCE partners think that by 2015, the total addressable market that Vblocks can address in private and public clouds will be around $85bn globally. ®