Is Virtualization Less Costly Than Desktops ?

In my field work, and in the press, there remains a question relating to whether virtualization is less costly as desktops in terms of the total cost of ownership or lifecycle costs. In this blog posting I will attempt to address this question.

As the case with all of my blogs, the content is mine and mine alone and does not represent that of my employer.

In many instances, I advise that for the business case assume that the economics are neutral; that there is no financial gain or loss on the TCO/ lifecycle side of the equation. If one assumes that this is the case, then the benefit side of virtualization will drive anticipated change. The direct answer seems to depend on who is asked the question. Many independent third parties suggest that the financial benefits are not significant, and that the benefits are more in the security and protection of loss (such as IP and consumer identity as examples). However, there remains a core following that advise that the benefits are in a very broad range of 30% or greater.

The answer really gets into the core of the TCO story board- what is the baseline that we are comparing virtualization to in the first place?

When you look at the TCO/lifecycle costs of the desktops there are a few things that we know right from the beginning.

1) If the dialog is about acqusition price of a desktop versus thin client, no contest. However, product is only about 8% of the overall TCO story.

2) TCO co-mingles direct and indirect expenses, so if the business case does not include end user productivity then part of the justification may be weakened

3) Energy management, once thought to be the home run in the virtualization picture is being challenged by those who believe that migration away from the desktop power consumption, simply moves the power consumption to the data center. I do not believe that this argument is solid since many businesses cannot separate a before and after scenario in manner to prove or disprove this logic. This to me seems more of an emotional response.

4) What burden does the IT infrastructure allocate for client computing? This potentially is a big point. If network, server, storage, and program management (again as examples) are allocated to the desktops as a part of the cost buld up, then the question arises, if virtualization is indeed a different model, then why continue such allocations. In this case, this may be a part of a compelling rationale.

5) The real benefits in my opinion can be articulated in a few operaitonal areas that a business can more easily put their arms around - patch management, configuration management, asset management, and technology refresh cycles. It seems to me that if a refresh cycle can legitimately be passed, say for example W7, and could be performed remotely for a percentage of the installed base, the benefits are significant. Virtualization as a by -product of the technology places a business at the best practice level for all of these operations as well as addresses the security issues discussed in the earlier paragraph of this blog. Deskside support as known today is not required in a virtualized environment.

User adoption may become an issue. Who would have thought that end users would become attached to older desktops as an emotional aspect of virtualizing? But yet, we do hear " I want my desktop" as an interesting refrain.

The back end infrastructure (the network readiness, the Active Directory, roaming profiles,and application stacks) remain as a part of the financial picture. I would flag that with W7, many businesses are already performing application rationalization. The effort of virtualize the applications could be an added step since we are already looking at the stack.

Virtualization should be viewed over a 5 to 7 year planning horizon. All of the one time only costs will be mitigated in that manner. Leveraging the 36 month lifecycle will result in "ugly" math. Also, scaling across a broad installed base will provide more seats to amortize the one time only costs.

The real financial benefits to virtualization may in fact be the enablement of BYOC and cloud computing. I cannot envision getting to these points without virtualization being mature and in place.

A long winded way of answering the question I know , but there is another point to be made. If a business is really weak in lifecycle management practices, the better the business case is for virtualization. It may be one of the few times in business where a weakness becomes a compelling driver.

Overall, virtualization costs less than desktops, however, "your mileage may vary...."