EMC's stake in VCE is nearly double what Cisco's is, but the fate of VCE ultimately may hinge on Cisco's evolving cloud strategy. Some Cisco insiders, in fact, say it isn't far-fetched to think that Cisco views VCE as little more than a footnote -- a stopover venture as it tries to control more and more of the customer cloud and data center spend and grow its Unified Computing System (UCS) into its next market-dominating business.
"Cisco is in a precarious position right now," said a longtime Cisco solution provider speaking on condition of anonymity, who said Cisco still looks vulnerable even after a year of painful restructuring. "They're watching smaller competitors write software and engage virtualized networking in a way that's going to replace Cisco's routing and switching, or at least make it so it doesn't have to necessarily be Cisco's. [Cisco] is trying to make the transitions they can but they're just not a growth company anymore. UCS is a hit, though, so they have to make that work for them now while they figure themselves out."
In 2008, a year before the UCS officially launched, Cisco convened a number of its key cloud- and data center-centric technology executives for a series of closed-door committee meetings. Presiding over those meetings as an executive-level sponsor was Padmasree Warrior, who had been named Cisco's chief technology officer in December 2007 after 23 years at Motorola.
The gathering included participation from well-known Cisco names like Doug Gourlay, the former vice president of data center solutions and now vice president of marketing at Cisco challenger Arista Networks; Jeff Spagnola, currently Cisco vice president of sales for cloud and managed services; and Rob Lloyd, now Cisco's president, development and sales.
The topic at hand: How would Cisco take what were then development-phase data center technologies and create a clear-cut road map for Cisco's role in the cloud?
What emerged from those meetings, according to several sources familiar with the discussions, was a four-phased plan. Phase one was to work with NetApp to develop a validated design -- the reference architecture that became FlexPod. Phase two included partnering with EMC, funding the VCE Coalition and marketing heavily around delivering a complete platform.
Firming up a cloud strategy behind UCS was a hugely important step for Cisco, which by April 2008 was deep into Project California -- its code-name for UCS – and that month would also acquire Nuova Systems, a Cisco spin-in venture majority-funded by the networking titan, whose core technology became Cisco's Nexus switches.
Warrior's charter from Cisco Chairman and CEO John Chambers at the time of her hiring was to do the firming up. In an interview with CRN at Cisco Partner Summit in San Diego in April of this year, she described Cisco's cloud strategy as the biggest accomplishment of her thus-far four-year tenure as Cisco's top technologist.
"I came in, and I started to put together our strategy for cloud, looking at the pieces we had and what we could do," she said. "We hadn't yet thought much about the role the network could play in cloud, so what I'm hugely proud of is the movement Cisco's made from having no name in cloud to being a clear leader in cloud."
Specific to UCS, Cisco will continue to build the ecosystem, Warrior said, and the developer partners. It has nearly 50 ISVs currently developing apps using the UCS API.
UCS has been a home run for Cisco, continuing to add customers and be an impressive revenue contributor for the vendor even during Cisco's challenging 2011.
In spring 2011, Cisco broke into the list of the top five server vendors in North America, and by the end of the year had passed Dell to become the No. 3 vendor of blade servers, according to IDC. In its most recent quarterly earnings, for the third quarter, Cisco touted that sales of the system grew 57 percent from last year while both of Cisco's two largest data center competitors experienced flat or negative growth in the same period. In August, Cisco won CRN's Annual Report Card award for Midrange Servers, sweeping the category for a major coup.
There's little question Cisco has made waves in the data center, and it has been growing UCS sales by whatever means necessary. VCE is just one conduit, a way to provide a "whole offer" data center architecture that, according to sources, represented phase three of the cloud plan Cisco hatched in 2008.
The next phase, however, is what is expected to push Cisco beyond relying on one or even two UCS-centric configurations and have it partner with a host of other vendors as it seeks to own more of the total data center play. A reference architecture launched by Cisco with Hitachi and VMware is one example of Cisco continuing to move beyond the Vblock and FlexPod models.
"It was not Cisco's plan to invest long-term in VCE," said a former Cisco employee with knowledge of the Cisco cloud committee's discussions. "They wanted to sell some Vblocks, prove it all works, then focus on their own architecture. Look at what Cisco has. They work very tight with VMware, they're certified with Microsoft, all that's really missing is the storage. I fully expect that VCE will fold eventually, the EMC people will just go back to EMC, and Cisco will do its own thing. EMC needs this to work a lot more than Cisco does."
Other Cisco executives say that Cisco and EMC were so caught up in making their convergence play known that the most important details -- the channel strategy, the internal systems -- were basically ignored.
"There was very little due diligence done about the go-to-market model behind VCE," said a former Cisco executive who was part of Cisco's VCE launch team. "It's very important to understand just how hastily this was launched. There was no real discussion of how it was supposed to get sold or how channel partners would make it work. It's just that the actual solution was so compelling, so the idea was to do the big splash and figure it out later. I don't think they anticipated how badly the internal politics would be a problem."
That said, Cisco insists it is committed to VCE. Cisco's Lloyd told CRN in April 2012 that Cisco would continue to fund the venture, and that it has "absolutely" been a success.
"We are extremely committed to the VCE model and extremely committed to the value we brought to customers," Lloyd said. "We think there's a whole runway to expand the reach of VCE and our capabilities much more through the data center partnerships we share. We are committed 100 percent and achieved exactly what we wanted to, which is a unique model that no one else has. You will see announcements from others beginning to emulate the model that VCE established, and that adds credibility to the fact that we did the right thing."
The former Cisco executive who was part of VCE's launch team said he isn't surprised Lloyd is cheerleading for VCE: Lloyd, as several other sources confirmed to CRN, was one of VCE's biggest internal supporters at Cisco and overruled other committee leaders within Cisco to invest in VCE as much as Cisco did.
"Rob Lloyd was a very key driver of this. Padma [Warrior] was involved, but this was all about some very significant opportunities for sales that Lloyd saw in the market," said the former Cisco executive. "I think Rob saw a clear opportunity to push on this -- and enough large deals out there to make this worthwhile. But on the Cisco side, the business leaders were not aligned to that vision. Many of us saw it as one of several opportunities for UCS instead of something to invest so much money in."
Lloyd pushed VCE forward almost despite some of Cisco's internal councils, even after concerns were raised by Cisco's data center board and its cloud board about whether so much attention on VCE and Cisco's EMC relationship was worth it. One Cisco source said it was Lloyd and executives who worked for him like Manjula Talreja, now Cisco vice president, global cloud business development, who shouted down committee naysayers to push VCE investment forward.
There's little question, according to sources, that EMC needed VCE to get off the ground much more than Cisco did.
"EMC was by far the most intense and invested in it," said the former Cisco executive. "Cisco had partial motivation for it, sure, but there were many people within Cisco's business units that had much more interest in working in open models with players like NetApp."
Below Lloyd's level at Cisco, support for VCE is mixed. Cisco's sales executives quietly mutter about how VCE sales have come in lower than initial expectations and that Cisco shouldn't have spent so much time and money being part of it.
Cisco analysts say it isn't surprising.
"Ninety-eight percent of all joint ventures of any type are essentially flawed," said Ehud Gelblum, managing director at Morgan Stanley. "I can't think of one that actually succeeded, all told. They haven't all been money-losing, though most of them do. This was probably a good idea at the beginning but it's been losing money ever since and creating conflict with the channels.
"Joint ventures just sound like a good way of uniting two companies together," Gelblum said. The answer could be as simple and as stupid as you do it because you're afraid of the other person infringing on your turf. EMC wouldn't partner with Juniper because it has this VCE thing set up. But then Cisco does partner with NetApp."
PUBLISHED OCT. 17, 2012
