The VDI Party is Over

When looking through some of the documents I have collected over the last years, I found an interesting statement made by Yankee Group. Their prediction was that VDI will get mainstream in 2009. Well, this prediction is a couple of years old and we all know that it didn’t happen. I don’t want to put Yankee Group on the spot; three years ago there were many other market analysts saying exactly the same. But what happened to the VDI hype? Attending Citrix Synergy in San Francisco and BriForum in London this month allowed me dig a little deeper into this.

For many years Citrix was the undisputed market leader in virtual desktop solutions. But about a year ago they started investing seriously into technologies that allow them to move away from VDI. The latest version of their strategy included things like the cloud management platform CloudStack, the unified storefront Cloud Gateway and project Avalon for delivering Windows apps and desktops as a cloud service. In essence, Citrix is moving towards aggregating and orchestrating. VDI is still an important building block, but obviously not the core foundation of future Citrix product offerings anymore. Citrix officially declared the post-PC era, just like VMware did. When I understand Mark Templeton’s keynote right, Citrix are concentrating on getting established as a major player in the cloud market. It looks like they want to be the ones who provide the platform and the infrastructure for delivering cloud services in a massively scalable way. Fair enough, but in such a world VDI is only a niche.

All this became even more obvious at BriForum. Brian Madden, Jack Madden and Gabe Knuth just released their new book called the VDI Delusion. Wait a second; is this the same Brian Madden who said only a couple of years ago that VDI will be ruling the world? In all fairness, when Brian made this statement he referred to VDI versus Terminal Services. Today it is clear that neither of these two technologies is the magic silver bullet. In fact, we are facing a growing diversity in application deployment methods – locally installed apps, streamed apps and desktops, remote apps and desktops, hypervisors on clients and servers, layering technologies and many more. Not to forget the also growing diversity of devices. It turns out that virtual desktops are not the ideal solution when using mobile devices without mouse, without keyboard and with small screens.

But even if the form factor of the endpoint is suitable for full Windows desktops, many users and some IT professionals are confronted with a “surprising” fact – VDI relies on graphics and multimedia remoting technologies. I personally don’t know what some people thought how graphical desktop and application elements were supposed to be delivered to the endpoint when using VDI, but I suspect they were hoping for some kind of magic after reading market analysis reports a couple of years ago. Now, after trying things out, they find out that even IT cannot beat the laws of nature, such as the speed of light resulting in network latencies. It turned out that the promised Champaign was only water. Bummer!

I still believe that remoting technologies have a great future, in particular after seeing very promising results of a number of RemoteFX tests I’ve done with the beta builds of Windows 8 and Windwos Server 2012. Not to forget Citrix HDX, Teradici/VMware PCoIP, Quest EOP or Ericom Blaze. But it’s remote Windows apps that people want to use when needed, but not full Windows desktops. The full desktop is not so important anymore, it’s the context-oriented delivery of apps that really matters – and most of them are running on Windows. Nope, the PC era is not over yet, but the VDI party definitely is.

3 comments

  1. What – a book by Brian Madden bashing something? I am shocked!

    Comment by Gordon Smith on June 1, 2012 at 3:41 pm

  2. Benny,

    I’ve been a huge fan since our days together as MVPs many years ago, but I do have two significant disagreements with this article. The first, and more trivial, is your claim that until about a year ago, “Citrix was the undisputed leader in virtual desktop solutions.” Because you make a distinction between VDI and Terminal Services, I assume you are speaking about XenDesktop vs. XenApp.

    “Undisputed leader” of course, can mean different things to different people, but the primary two areas that I think most would agree constitute leadership would be market share and product superiority. Regarding market share, the studies that I’ve seen all point to VMware as either leading or tied with Citrix. The IDC Worldwide Desktop Virtualization 2011 Vendor Analysis (available on Citrix’ Web site), for example, shows equivalent market shares between the two. Morgan Stanley reports show that while Citrix is expected to catch up with and pass VMware in the next year or two, VMware has led up until now. In either case, it would be hard to argue Citrix as the “undisputed” market share leader.

    From a technology standpoint, XenDesktop and VMware View both have their pros and cons. I think that a VMware advocate could argue just as passionately and convincingly as a Citrix proponent. The organizational objectives, environmental factors, staff experience, etc. all play a huge role into which platform is optimal. Again, though, I think there is zero justification for claiming Citrix as the undisputed leader.

    My more serious disagreement comes with the central tenet of your post which is that “The VDI Party is Over”. You mention Madden’s book, “The VDI Delusion” I recently wrote a responding blog post titled, The VDI Delusion illusion http://www.bythebell.com/2012/05/the-vdi-delusion-illusion.html that explains why the VDI party is really just getting started. I won’t recap my whole post here, but the combination of Moore’s Law and rapid innovation along with BYOD pressures and cloud computing is leading to an acceleration in VDI deployments.

    While I have not seen any recent aggregated industry statistics around VDI sales, based upon the business into which I do have purview, I would say that my contention should be “undisputed”.

    Comment by Steve Kaplan (@ROIdude) on June 2, 2012 at 6:28 pm

  3. Benny,

    I would continue to fill your head with ego growing comments however we all know what a smart fellow you are and your contributions to the market that is growing and will come.

    With that being said I have to also add my disagreement to Steve’s adding that the one big thing missing from the aforementioned book and this blog is time. I’m consistently amazed at how we declare defeat on a buzzword or phrase like VDI and perhaps we are right, but time will deliver the true solution and a distributed desktop technology when it delivers a superior desktop experience with more features then the traditional paradigm will prevail over time in my opinion.

    I think the major disconnect is that we are still evolving from Terminal Services solutions which were great and served the purpose most talk about VDI usage for (purpose built virtual desktops for task workers). Whereas there are a number of great technologies which are being developed to take advantage of any endpoint to deliver the relevant solution end to end which all can benefit from all around.

    I’m excited about what is coming and perhaps you can declare ding dong the witch is dead, however what happens to OZ now? Great stuff is coming.

    Comment by Peter Bookman on June 4, 2012 at 4:46 pm

Trackbacks

  1. VDI feest voorbij? Was het al begonnen dan? « EarlyBert
  2. VDI, ok ? What’s next ? | archy.net
  3. Virtual Desktop and Applications mean less tham Data | JasonConger.com

The comments are closed.