Wednesday, May 19, 2010

VDI can save you big! Don't count the savings just yet.

Many solution providers are pitching VDI as the next big cost savings initiative for frugal IT executives to implement. However, the savings may not be all they are promised to be.

In a research study Alinean participated in on the total lifecycle costs of VDI compared to traditional desktops, VDI has proven to be about 10% more expensive in certain mainstream scenarios - particularly for traditional Office Workers, in an environment that is already well managed.

The objectives in the study were to understand customer perceptions and expectations from VDI, gather real world data from deployed organizations, and
provide customers with TCO guidance from a collection of industry experts. Although the study was sponsored by Microsoft, I can tell you that having participated in the debates and research that the results are factual and balanced.

First, surveys of 105 organizations in Financial Servies, Government, Manufacturing and Retail,indicated that unanimous market perception / sentiment is that VDI has a lower TCO than the PC.

Deeper analysis of actual costs and savings indicate that VDI is indeed cheaper than the PC in certain use cases / business applications, for example delivering VDI to structured task workers. However, the results also clearly indicated that VDI can be more expensive for office / knowledge workers, particularly when the organization's PCs are already well managed. Because a well managed environment is often a prerequisite to implementing VDI, this represents an interesting value realization connundrum.

For the study Total Cost of Ownership (TCO) was used to compare the costs of VDI vs. traditional PCs, tallying not just the up front costs, but the total costs over the entire lifecycle from planning, through deployment, management, evolution and retirement. Created by my friend Bill Kirwin at Gartner in 1989, it has been used since to great effect to help buyers make economic decisions for key IT solutions.

The TCO model compared the Direct Costs of both solutions, including CAPEX – Hardware, Software, and OPEX – Administration, Operations, Fees. As well, the service levels and capabilities of the systems need to be considered in any good TCO model, tallying the Indirect Costs for End-User Operations. The indirect costs include service level issues such as downtime / accessibility / support wait times when issues occur, as well as additional time users spend solving their own issues, or futzing with systems to get them to do what they need it to do.

According to the research study, VDI does indeed provide benefits in many situations including:
> Centralized Management - manage physical and virtual clients from a single console, centralized desktop lifecycle management
> Enhanced Security and Compliance - data is always locked in the data center, improved compliance through centralization
> Anywhere Access for Connected Devices - access desktops from any connected device, enable rich desktop experiences on thin clients and older PCs
> Increased busienss continuity - datacenter grade business continuity and disaster recovery, quickler resolution to desktop failures.


However, the bottom-line was surprising to many who beleive that VDI always delivers a lower TCO than traditional desktops.

Comparing the costs for VDI vs. PCs in a well managed office worker environment:

1) VDI was 9% more expensive than Windows XP SP3 -> $908/client/year for VDI versus $836/PC/year for Windows XP.

2) VDI was 11% more expensive than Windows 7

Of course these figures depend on the particular application and organizations opportunities for improvement, but are eye opening all the same.

The study considered environments where VDI may not be a panacea of savings, in particular supporting office / knowledge workers, and in an already well managed environment (Rationalized according to Microsoft's IO maturity scale). Many are considering broad VDI deployments, with office / knowledge workers the majority of the population in most organizations, so this scenario is mainstream. In researching organizations that have implemented VDI for office workers, issues were prevelant with user experience, This loss of productivity added to the indirect costs for VDI vs. PCs. The study indicates that many office worker applications, like VoIP, might have degraded performance, and graphic intensive applications currently don't always perform well over WANs. Start-up delays, performance issues, availability issues and offline access problems add to the UX issues.


Also, because implementing VDI often requires a certain base capability in order to deal with the added data center complexity, organizations often need to advance management capability / maturity prior to implementing VDI. Because Rationalized environments are already efficiency and optimized, VDI often cannot drive enough additional labor savings to overcome the added capital for VDI datacenter hardware and rather expensive software licensing investment.

For environments that are unmanaged or not well managed, the savings from VDI for office workers turn positive, but in these environments the capability to effectively implement VDI and acheive the savings is diminished. For task worker environments, the savings for VDI deliver between 10-30% (source: IDC), but can these savings be realized and the solution effectively implemented? And when the UX issues are considered and indirect costs are compared, do the savings still remain?

Having participated as an industry expert to review and validate the results, the study is an important one to open perceptions towards VDI. VDI indeed does not offer guarenteed savings. Certain applications are perfect for VDI, such as virtual call centers and global sourcing, but other mainstream applications are not cost effective at this time. As costs drop and UX improves, this can change, but for now, mainstream VDI will not deliver cost savings for most organizations.

Therefore, for today, the application is the most important driver for VDI, not cost savings, and this study should maintain decision makers focus on that fact.

A summary white paper on this important research study can be found at: www.microsoft.com/vdi


To calculate your own unique potential for savings, use the following on-line tool (registration required):
https://roianalyst.alinean.com/microsoft/virtualization/

Microsoft blog on VDI TCO can be found at: http://windowsteamblog.com/blogs/business/archive/2010/05/10/simplifying-desktop-virtualization.aspx

No comments: