I'm a fan of virtualization in all its forms, but there are some issues to consider on the desktop side of the house.
VDI-Virtual Desktop Infrastructure-is the latest buzzword in the already buzzword-heavy world of virtualization. You already know that virtualization can work all sorts of miracles such as reducing power and cooling costs, lowering hardware cost, utilizing hardware better and, with the release of Hyper-V, creating world peace. Ok, maybe that last one is asking a bit much…
Personally, I’m a huge fan of virtualization in all its forms. After all, why buy eight servers when one will do? Why power and cool eight servers when one will do? Since the typical x86 single-purpose server is only 10% utilized, why waste 90%+ of a capital investment? I don’t know of many organizations that would build a 1,000 unit office building, but only fill it with 100 people! You get the point.
On the desktop side of the house, the virtualization equation is a little different and a bit more of a challenge, but I think it could be a major boon. Although many organizations have standardized desktops, desktop software and desktop management techniques, there are still the outliers that have not.
On the ROI side, which is generally very easy to demonstrate on a server consolidation/virtualization undertaking, hard dollars for VDI may be a bit more difficult to quantify. Part of the idea behind VDI is that you can have a lot of the benefit of a terminal services-type environment without the shared resource hassles that come in to play with Terminal Services and Citrix. This could mean that a user’s desktop PC could be replaced with a less expensive, less power-hungry thin client. But users have grown accustomed to, for example, having local devices like CD ROM drives and a good way to stream music from the Internet. But, are they "business essential" services? Sure, in some places, a CD ROM drive is still a necessity, but that is becoming more and more rare.
If you look at the desktop PC as nothing more than a productivity tool and not something to be used for employee entertainment, and you’re able to provide either the same or a higher level of service while possibly saving a few dollars, there is a potential value proposition for virtualizing the desktop infrastructure, especially when the VDI solution is accompanied by good broker software and application virtualization, which further decouples the software from the hardware. Imagine this scenario: A user’s complete Windows XP-based desktop is delivered to her over the network with the user equipped with just a thin client. The applications she uses are delivered to the virtual XP host using an application virtualization technique. Upgrading applications becomes a breeze by virtue of the application virtualization software and desktop deployment becomes a trivial matter due to VDI.
With VDI, you can also begin to provide better anytime, anywhere computing for your users. Even better, because users would be using the exact same desktop wherever they go, all potentially sensitive data stays on your servers and never has to really traverse the Internet. So, there is possible long-term cost savings, possible better service and a definite increase in the level of data security.
Of course, VDI will not be for everyone. For users with high processing needs, VDI is a non-starter. If, however, you’d still like the advantages of desktop centralization, you can look at some of the many centralized desktop solutions that are available on the market.
Personally, I see VDI eventually taking off in such as way as to make today’s virtualization projects look small in comparison. We will be investigating this technology at Westminster College in the coming months and I’ll report back on our progress.