Are we there yet? No it is not the annoying chant of the kids in the backseat but rather the question that is being asked about the maturity and readiness of virtual desktop or virtual desktop infrastructure (VDI).
For those who have escaped the hype around the newest trend that is poised to revolutionize computing let me give you the generic definition.
VDI is hosting a desktop operating system within a virtual machine (VM) typically running on a centralized server or data center. The term was coined by VMware Inc.
Now let’s take this virtualization model even more into the stratosphere (noticed I did not use the term cloud). There is a rapidly emerging market where data centers or cloud (sorry) outsourcers provide hosted virtual desktop services. According to a report by Gartner, hosted services accounted for more than a half million desktop units as of March 2009, but will grow to 49 million desktop units by 2013, and may make up as much as 40% of the worldwide “professional PC market” by revenue.
Now I can hear the collective reality check of the fellow old timers “wait a second this is taking us back to the days of the dumb terminals and monolithic computing that we were told was bad and client/server was the future”. Yes I am among the skeptics that question the validity of this back-to-the-future logic as well. But when you separate the hype from the benefits that are based on reality you can see why this model should not be ignored.
I will attempt to summarize the benefits into two categories for illustration purposes.
One benefit is the reduced cost associated with provisioning desktop hardware. This benefit alone is usually the main driver that supports looking at VDI for the large enterprise that has a consistent demand for deployment due to attrition, M&A, and organic growth of the enterprise.
Let’s take a hypothetical financial institution that has a workforce of 100K. With attrition associated with consultants and full time staff of a conservative 1% per month this equates to a thousand devices that require to be physically managed globally on a monthly basis. Let’s apply a conservative range of 2-4 hours per device at a loaded cost of 50 dollars/hour. In pure operating expense this could easily exceed 2 million dollars per year. Those type of dollars usually can get the attention of the C level.
The other potential benefit with the VDI model is the possibility of extending the lifecyle of the desktop hardware. Again using our hypothetical financial institution with 100K desktop devices with a 3 year refresh lifecycle, deferring the refresh to 4 years has the potential to saving millions in capital expenses.
You might be asking how can VDI extend the life of you current workstations. I used the term “dumb terminal” previously but the more political correct term of today is “thin-client”. The thin-client workstation needs only enough resources to run the remote controlling software as part of the virtualization solution. Users of such a thin-client may not even know that their software is actually running on a centralized computer. The need to replace workstations due to outdated capabilities such as the OS, memory, storage, or CPU are reduced with the thin-client deployment available with the VDI model
Of course we don’t often have 100K devices to use in our business case but I have clients with under 6K devices and with a much higher attrition rate due to the heavy use of consultants so they are challenged with managing the provisioning of 200-300 devices per month. The numbers are still pretty compelling.
So are we there yet? The technology is readily available by multiple proven providers. I sell concepts not products so I won’t mention specific product vendors in this article. There are real life case studies and success stories that can be found. In my opinion the desktop virtualization is well on it’s way.