If virtual desktops great, why not used more?
Desktop virtualization skeptic calls supporters on not using VDI more prominently themselves. But from user perspective, pure VDI isn't goal--flexibility, efficiency are.
Virtualization analyst Brian Madden asks an excellent question:
If VDI (virtual desktop infrastructure) is so great, then why aren't you using it?
It's a really good question, isn't it? Brian observes that however not in our backyard (NIMBY). Brian, a bit of a virtual desktop skeptic, marks his words: "VDI is a niche. 10 percent max at best." He does give himself some wiggle room, saying that's "10 percent of all users, or 100 percent of users for 10 percent of their apps." But either way, a niche., and however enthused we may be about extending the wins of server virtualization over into the desktop realm, we, personally, are not using desktop virtualization. You don't see analysts and developers doing so. And even the folks you meet from Citrix, Microsoft, Quest, VMware, and Wyse--the people selling VDI, for goodness' sake!--use traditional "fat" notebooks. The implication is that we may be pro virtual desktops, but
I'm a bit less skeptical. Here's why:
Maturity - There have been an awful lot of technologies that labored years to achieve a level of capability, completeness, and maturity needed for them to really take hold. The Internet, for example. You know--that thing most of us now don't know how to live without? Inter-networking was born in the early 1970s, but it wasn't until about 1995 that it really took off. When it sparked, the Internet blasted from totally geeky to part of our social and business fabric within three years. It was an overnight success--it just needed 25 years to get there. SMS texting, Facebook, and oh yeah, x86 server virtualization were much the same. I sense a similar "no, no, not really, no, enh, maybe, yes!" progression with virtual desktops. Many of the prerequisites and requirements simply were not met in years past. But now we routinely have high-capacity networks, excellent server consolidation, multimedia-capable VDI technologies, etcetera--these make it a whole different ball game. VDI 2011 is worlds beyond VDI 2008, 2005, or other previous incarnations.
The Right Tool for the Job - Mobile users are not the best-use case for desktop virtualization. Unlike those who work in a single office (or on a single campus), we can't be sure of having always-on, low-latency networks. So we tend to take our IT lives with us, on laptops. Salespeople, analysts, those who attend a lot of conferences--if you're looking to us for confirmation that virtual desktops are a good idea, you could have virtual desktops blanket corporations across the world, yet still see low levels of penetration in the extraordinarily mobile. We're trailing indicators.
It's also worth noting that I do use GoToMyPC, RDP, and VNC remote desktops--including from the road. They're very useful for accessing apps that live only in one place, as well as for system management and software development. You won't see me using VDI as my one-and-only from the road, but it is part of the mix.
Brian seems a fan of client-side virtualization that lets mobile users check out their VMs to their laptops. Me too, though I think it's not quite as well-developed or mainstream yet. In order to get the easier, larger-scale economic wins that get virtual desktop projects funded and approved, most enterprises need to focus on use cases that aren't the hardest, and that cover large user populations. The most mobile users generally aren't the best candidates.
The Cloud - Many people do not view Software as a Service (SaaS), cloud computing, or rich Internet applications (RIAs) as part of the desktop virtualization landscape. I think that's a tremendous mistake. Sure, SaaS and VDI tend to use very different technical underpinnings--Web 2.0, not classic Windows desktop stuff. But from an enterprise's point of view, Web apps address many of the same problems: consolidating back-end and support requirements, reducing the cost of serving users, making deployments and updates more flexible, and providing a more scalable way of delivering apps.
Most users now either do, or could, spend much of their working lives in the context of a Web browser. Writing documents, fiddling spreadsheets, editing presentations, handling e-mails, filling accounting forms, ordering supplies, learning about the competition, collaborating with co-workers, meeting with partners and prospects--I increasingly do all of these things online. As a content producer, I still spend a good deal of heads-down time in heavyweight apps, but less every year. There's much less of my desktop to be virtualized because it's been so extensively Webized.
Brian's right. Pure, unalloyed VDI will be taken up by only a fraction of the desktop market. But pure VDI alone isn't the only thing on the table. IT can pick and choose several different flavors of desktop virtualization, combining them with complementary approaches like application virtualization (e.g., App-V) and Web apps. They're all aimed at a more consolidated, efficient, flexible way of providing IT service--and they're all welcome in my backyard.