The Year Of VDI: Never

The enterprise popularity of tablets and smartphones at the expense of PCs and other desktop devices is also sinking desktop virtualization.

The heyday of virtual desktop infrastructure came around 2008 to 2010, as companies sought to cut their desktop computing costs -- VDI promised savings from 10% to as much as 40%. Those savings were possible despite the additional engineering and server investments required to implement the VDI stack. Some companies even anticipated replacing up to 90% of their PCs with VDI alternatives.

But something happened on the way to VDI dominance. Employee demand for mobile devices, in line with the BYOD phenomenon, has refocused IT shops on delivering mobile device management capabilities, not VDI. On-the-go employees are gravitating toward new lightweight laptops, a variety of tablets and other non-desktop innovations that aren't VDI-friendly. Mobile employees want to use multiple devices; they don't want to be tied down to a single VDI-based interface.

Given that the VDI interface is at best cumbersome on a touch platform, there will be less and less demand for VDI as the way to interconnect. Highly mobile alternatives will only increase in popularity over the next few years as the war between Apple, Android and Microsoft/Nokia intensifies and they produce better and cheaper products.

Meantime, PC (both desktop and laptop) prices will decline even faster as the industry tries to sell its overcapacity. Already, Dell and Lenovo are lowering prices just to hold their volumes steady. And with even more devices entering the market (smart TVs, smart game stations, etc.), the market will get bloody. The end result for IT shops will be $300 laptops that are pretty slick and come loaded with Windows (maybe even Office).

At those prices, VDI will have minimal or no cost advantages, especially when taking into account the backend engineering costs. If companies can buy $300 laptops or tablets that are preferred by most employees, they'll be hard pressed to pass them up and impose VDI. By the end of next year, VDI solutions could cost more than conventional client devices (e.g., that $300 laptop). That's because the cost of VDI's main components -- servers, engineering work and support -- won't drop nearly as quickly as PC prices.

Global CIOs: A Site Just For You Visit InformationWeek's Global CIO -- our online community and information resource for CIOs operating in the global economy.

There's no escaping the additional engineering time and attention VDI requires. Its complex stack (either Citrix or VMware) still requires more engineering than a conventional desktop solution. And with that complexity, expect bugs between the client, VDI and server layers to impact user experience.

At Allstate, for example, we've experienced more than our fair share of defects in our recent enterprise upgrade of both desktops/laptops and VDI across 40,000 clients. Despite an updated and fully engineered VDI solution, we continue to find bugs between the virtualization layer, Windows and third-party products. And this is for what should be a mature technology platform by now.

Faced with higher costs, the need for more engineering resources (which are scarce) and employee demand for the latest mobile devices, organizations with extensive VDI deployments will allow significant erosion of them. Some will reduce the scope of current VDI deployments. Others evaluating VDI will jump instead to mobile-only alternatives, more focused on tablets and smartphones. IT shops will not want to support VDI for an employee who also has a tablet and/or conventional laptop or desktop environment because that setup essentially doubles the cost. This is a long fall from the lofty goals of 90% VDI deployment.

An interesting phenomenon in the rapidly changing technology world is when a tech wave gets overtaken well before it peaks. Think optical disk storage in the data center in the 1990s. Or, more recently, netbooks, whose cost and simplicity were overwhelmed by smartphones (from below) and Ultrabooks (from above). It's very difficult for any technology to carve out a sustainable market niche based on cost advantages alone.

I'm sure there will still be niche VDI applications, such as in offshore development centers, especially where VDI enables better control of software licensing. And small segments of the user population will swear by the flexibility to access a client device from anywhere they can log in, without having to carry a device.

Long term, however, VDI solutions will serve maybe 10% of the employee population, perhaps 20%, but not more.

What is your company's experience with VDI? Where do you see its future? Please tell us in the Comments section below.

Learn more about virtualization at the Interop conference and expo in New York, Sept. 30 to Oct. 4.