There's been an increasing amount of buzz lately around the fact that Windows 8 will be shipping with integrated Hyper-V (or at least features of Hyper-V). Nerds and admins and such (Tim, Benny, et al.) will be happy to hear that they don't need to get VMware Workstation, VirtualBox, or Parallels, but what does this mean for the bulk of the corporate desktop users?
As it stands, 99% of them won't give a damn. It will be like Media Center, which I'd imagine most people don't know they have as part of Windows 7 Home Premium (or the other versions it comes on). That said, if Microsoft and third party companies do things correctly, maybe the client hypervisor will get a second wind.
When we talk to people about client hypervisors (most recently on an episode of BG Live with Brian Gammage from VMware), we hear about the fact that users don't want/need/care to know about whether or not they're using a client hypervisor. In order for a product to gain widespread adoption, this is one of the things that needs to be addressed and solved elegantly, but there so much that goes into a truly seamless solution that it makes it almost impossible (even the optimists will say it's very difficult).
It's easy to argue that the best way to give a user two desktops is to give them two physical devices, which is the only way you can get the ultimate mixture of isolation and usability. Of course, that's far from seamless, and it's not very helpful to the client hypervisor makers, either. They've spent the last four years trying to arrive at a solution that is both secure and flexible with a decent user experience, all while supporting a massive amount of hardware. For these companies (and I'm talking about Citrix, Virtual Computer, MokaFive, and Virtual Bridges, for the most part), most of their efforts have been on supporting the hardware that their clients use. It's like paying bills. You earn X dollars per month, and of that money, you need to pay rent, utilities, groceries, and so on. Most people pay the rent, then figure out how to manage the rest of the costs. The same is true with client hypervisors - most effort is spent just getting it to work, and whatever is left is spent on adding new features.
Building a hypervisor into Windows 8 with all the awareness of the client's hardware (power management, network connections, resource info, etc…) means that Microsoft will have assumed a huge chunk of the development budget, so to speak, which frees up the current client hypervisor companies to focus on things like:
- seamless integration
- native-like performance
- native hardware access
- enterprise management (remote wipe, provisioning, etc...)
- out of band management
- single disk image
- layering (gasp!)
- user environment portability
Of course, it also opens the door for other companies to bring in their own solutions now that they don't have to worry about the hypervisor components, but companies like Virtual Computer, who probably have the most complete solution available today, have a leg up on management since they've been doing it for so long. Consider Citrix, too, which already has a client hypervisor solution in XenClient. They're only now getting past the toddler phase with it, so they too could benefit from no longer having to worry about the hypervisor. It would give them more time to focus on the things that would make XenClient great: portable VMs, offline VDI, and security. Don't forget about Quest, either, who've hitched their wagon to Hyper-V like no other platform. With Hyper-V in the datacenter and on the desktop, not to mention the layering capabilities that they have access to from MokaFive, they could be in a great position to put client-side Hyper-V to use early.
All that, though, doesn't amount to a hill of beans if Microsoft doesn't make a built-in client hypervisor accessible to the third parties. It remains to be seen exactly how the Hyper-V components will be implemented (could MinWin be involved, or is Hyper-V too high up the stack for that?), but the more open they are, the more third parties can develop for it. With vast hardware support off the table, the third parties can now focus on more perfecting the technology that we all once thought had so much promise.