At this month’s New England VMware User Group meeting, an attendee told me about a conversation he’d had with an EMC higher-up. The topic turned to VMware, and the EMC executive said something like, “We thought it would be a good investment, but we had no idea it would turn into this!”
Back in 2003, when the acquisition happened, VMware was only four years old, and server virtualization was still a fringe technology. Just check out this CNET story on EMC’s VMware acquisition: It describes VMware as “a start-up that sells software to make servers more flexible” and doesn’t even mention the word “virtualization” until the fifth paragraph.
Now virtualization is one of the hottest IT markets, and VMware is its leader, with nearly $3 billion in annual revenue and a yearly conference that drew 14,000 people. So why would EMC want to get rid of VMware? Or why would VMware want to get rid of its core virtualization business?
That’s what two observers have suggested in recent blog posts.
(Wardley, a former executive at open source vendor Canonical, also writes that “a dominant position in the enterprise is no guarantee for future success” and cites Novell’s demise as an example. But everybody was talking about that when Hyper-V came out three years ago, and it hasn’t exactly proven to be true in the virtualization market.)
Sure, KVM has drawn some interest from the open source community, but VMware users aren’t ditching vSphere and flocking to Red Hat Enterprise Virtualization en masse. And virtualization is an underlying component of cloud infrastructures, among organizations that build private clouds and providers that host public clouds. Wardley says it’s a “misconceived” notion that virtualization is necessary for cloud computing, and while he’s technically right, the cloud market is by and large relying on virtualization.
Even if you agree with Wardley and think open source virtualization and non-virtualized cloud computing will take down VMware’s core business, there’s still the issue of timing. As GigaOm’s Derrick Harris wrote in response to Wardley’s blog post:
VMware itself acknowledges that platforms are the future — thus its heavy activity building out its SpringSource business — but the last time I asked someone from VMware what the timeline for such a transition is, I heard it would be about 10 years. … And with estimates that server virtualization isn’t really close to peaking in terms of adoption, there’s still a lot of money for VMware to make.
So let’s just say VMware is holding on to its virtualization business for the foreseeable future. There’s still the question of whether EMC will hold on to VMware. One person who sees the possibility of a sale is Software Advice consultant Christopher Baum, who has identified VMware as a potential Hewlett-Packard acquisition target. He writes:
In the grey area between glass rooms and cloud computing is virtualization. VMWare supplies software that makes a single physical server appear to be several virtual servers. Virtualization can increase security and ease management for individual applications or user classes. The firm would be a good candidate for HP this year.
Nothing Baum says is incorrect, but it’s definitely not complete. VMware would be a good candidate for lots of vendors to acquire — if it were for sale, which it’s not. Any server, storage or infrastructure vendor would love to own the most dominant company in the virtualization market.
I’m not saying things can’t eventually change. The Roman Empire fell, General Motors filed for bankruptcy and Foreigner went from selling out arenas to playing at VMworld. One day a time will come when virtualization isn’t a hot market and VMware isn’t the top dog. But until that time comes, EMC isn’t selling VMware, and VMware isn’t selling its virtualization business.