Why VDI use will never hit 100%. (And why that's ok.)

"What?" you're probably asking, "Brian hasn't written this article about ten times already?

"What?" you're probably asking, "Brian hasn't written this article about ten times already?" Yeah, it was a surprise to me too.

As a bit of background, I was talking to someone on the phone yesterday who asked, "I know you think VDI isn't going to ever have 100% penetration, but I wonder if you can explain exactly why not?" I told him that I'd send him a link to an article on that, assuming I had written it already.

Well... turns out I hadn't. (Until now.) So while this might be old news to regular readers of the site, today I'd like to focus on why VDI is never going to hit 100%.

First, we have to define VDI. I typically describe VDI as a Microsoft Windows desktop running as a virtual machine in a location other than the on the device (laptop, desktop, thin client) on the user's desk. My definition is about technology, not business structure, so my definition of VDI also includes DaaS.

Closely related to VDI is Remote Desktop Session Host (RDSH) which is where a user gets a slice of a Windows VM instead of the whole VM. Some people also include remote physical workstations (like HP Moonshot or blade PCs) in their definition of VDI, and others include single-user instances of Windows Server.

The point is that when someone asks about the penetration of VDI, you have to ask whether they're also including RDSH? Are they including DaaS? Are they including remote 1-to-1 physical machines? (Obviously the broader your definition, the higher the adoption rate will be, so you can use your "How to Lie with Cost Models" knowledge to tilt the answer however you want.

For me personally, the important aspects of my definition of VDI are (1) VDI is about Microsoft Windows, and (2) VDI involves users remotely accessing those Windows desktops (regardless of the form factor). So right off the bat you can see that VDI will never hit 100% of end users because some users don't (or soon won't) need Windows desktop applications. (A small business or startup with only Macs and web apps doesn't need VDI.)

Okay, so what if we change our question to ask "Will VDI hit 100% penetration when it comes to Windows desktop applications?"

That's actually a bit more plausible and seems to make logical sense in the context of the future of computing. Sure, a lot of Windows desktop applications are being replaced with cloud, SaaS, web, and/or native mobile apps, but the reality is that there are still a lot of existing Windows desktop applications out there. And let's be honest—if you have a Windows desktop-based line-of-business application in 2014 that you haven't migrated to web, Java, iOS, SaaS, cloud, or mobile yet, then there's probably a good reason for it <cough> complexity! </cough>, so it's likely that it's going to be a Windows desktop application for a long time.

In those cases you could see how VDI (or DaaS or RDSH or remote server) usage could actually dramatically increase over the next few years as customers are forced to deliver and support the occasional "old" Windows desktops application here and there without actually wanting to deliver and support full Microsoft Windows operating systems on every end point.

Then again, VDI is still about remote access to those Windows applications, which means your users have to be online and have a decent internet connection. (So that right there is another reason we're not going to see 100% VDI usage.)

But if you boil down everything else, the main reason we'll never hit 100% VDI use is because not everyone is using VDI today. So VDI really has to "compete" with the status quo of just installing apps locally on desktops and laptops. And especially as Windows desktop applications become fewer, companies aren't going to be too exited to go out and buy, rent, build, design, and implement all this fancy new VDI technology to support what they view as dying apps. So they'll go out and get the latest cloud, SaaS, web, and mobile apps for as much as they can, and the Windows apps will sit on a nicotine-stained CRT-based desktop in the corner for the rest of their days.

(By the way, if you're curious about the future of the Windows desktop (since it clearly won't be VDI), check out my post from last December, Brian Madden's vision for the future of the Microsoft Windows desktop, boiled down to a few simple bullets.)

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

Brian, Did you add an extra zero? I don't think VDI will reach 10%. And that's also ok!


Amitabh is spot on. According to Gartner VDI use will be less then 8% in 2017. Given the decline in the use of legacy Windows applications it may never go past 8%.


Here is how these predictions are made


Instead use your own common sense. Your desktops use cases are here to stay for a long long time. The only thing that is likely to be true is that you won't have one type of desktop. I.E Physical, VDI RDSH, and increasingly diverse apps. All the rest does not mean much. In real life it means more complexity and cost that needs to be handled.


windows is such a poorly debugged series of device drivers - why any of you waaste your time worrying about how it is delivered is really quite giggle inducing