Why the Hottest New Tech Solution Will Go Cold
Virtual Desktop Infrastructure could become a big market, but the challenges will bring it down first.
Virtual desktop infrastructure (VDI) is one of the hottest new technology solutions in the market today, enabled by the revolution that is server virtualization. Many vendors (and investors) are hyping VDI as the answer to the thorny problem of managing and securing desktop/laptop computers. And with Windows 7 (MSFT) as a catalyst, it's supposed to grow into a billion dollar business within a couple of short years.
So says the hype.
We believe that while desktop virtualization is an important trend, the VDI hype has gotten far ahead of IT market realities, and is likely to steer many down the rabbit hole with Alice.
A quick review: Server virtualization technology (popularized by VMware's (VMW) ESX and vSphere) separates the physical server from the operating system and applications. Up until this year, the main driver of server virtualization has been server consolidation; that is, running multiple virtual applications on a single physical server -- essentially mimicking the operation of many servers on a single one.
The second major driver of server virtualization is mobile workloads. This allows application packages (called virtual machines) to move from computer to computer while continuing to serve users. This capability enables both inexpensive clustering and disaster recovery, and is a primary technology driving the "Cloud" -- it pools computing and storage resources for better utilization and performance.
VDI is a derivative of server virtualization. It's a desktop operating system and application(s) running in a server virtual machine, which then "remotes" graphics (or screens) to a desktop/laptop computer connected to the network.
It's very similar to a solution that's been around since the mid-1990s, generally called "Terminal Services (TS)," the best example of which is Citrix Systems' (CTXS) XenApp. If you use Bloomberg Professional, you're using XenApp.
The major technological advantage of VDI over TS is better application compatibility. However, VDI is by no means perfect, and I'd like to share a few of the downsides I believe will limit the use of the technology for the intermediate term.
From an investor perspective, the VDI market is projected to be much larger than the market for server virtualization. The thinking is, "There are 30 million servers and 300 million to 400 million desktops… Therefore, VDI has to be a big market." Hence the hefty earnings multiples given market leaders Citrix Systems and VMware.
Investors are expecting both vendors to individually produce somewhere between $100 million and $150 million in 2010 VDI revenue -- essentially two to three times 2009 revenue.
While stoking the fire, neither company will actually break out VDI revenue, so it's going to be hard to ascertain the real numbers. Nevertheless, the revenue growth is still in 2010 analyst expectations.
Marker Advisors has been following the virtualization world for more than six years, has participated in nearly every VMWorld conference worldwide, and has tracked the VDI market from its infancy. We at Marker Advisors speak weekly with customers and partners that are testing and deploying VDI technology all over the world. After conversations with many dozens of professionals working with this technology, Marker believes the following to be the actual state of the VDI market:
1. Most VDI proof of concepts are failing. Many companies are testing VDI using proof of concepts, and our analysis leads us to believe that most are failing, with the customer instead continuing to use existing technology solutions (most often TS). The reasons given include
- The high cost of hardware and software infrastructure to run VDI. From the IT perspective, customers must invest in new data center server clusters, storage, and software to provide the equivalent of the total desktop computer power they want to serve. In fact, customers must replace cheap desktop computing and storage with expensive data center computing cycles. This is a major upfront capital expenditure that is difficult to justify in a tough economy.
- Users see little difference when comparing VDI to the TS application they are already running. From the user's perspective, if they've needed remote capabilities, they've been running XenApp or Microsoft's TS -- likely without even knowing which. With the current state of VDI technology, there is no end-user benefit. Let's put that another way: After all the money spent by IT to deploy VDI, the user sees no improvement in his computing experience, and often sees a degradation.
- Vendors' lack of focus on the user experience, or "Why a user would want their desktop or applications delivered this way?", is a major obstacle to widespread deployment. In fact, the vendor messages today are all IT-oriented. "More secure." "Lower cost of ownership." "Easier to provision and maintain." Hardly areas that concern a knowledge worker trying to get his job done. Selling just to IT is fine when it's only servers being discussed, but users use desktops. The selling of VDI inevitably incorporates business units (users), which vendors aren't sufficiently addressing -- in terms of product and sale.
2. "Use Case" Solution. The consequence of the above is that in practice, VDI is being deployed in only a few specific use cases, particularly where it is either replacing a dumb terminal-like experience, or where users aren't accustomed to having a say in their user experience.
- Security and Compliance. Companies with offshore and/or onshore contractors are using VDI to deliver secure desktop applications for temporary use. Or in high-security situations, secure desktops for specific applications.
- Fast Provisioning. Education organizations are using VDI because they can quickly provide an education desktop to a user base that is by nature very transitory.
- Mobile Workforce, Fixed Computers. Health care and education organizations are enabling workers/students to walk up to any computer, log in, and have their desktop follow them to that station.
Clearly, there is a market today for VDI; however, not a billion dollar market. To make things more difficult, there are four public companies that are struggling to get a foothold: VMware (View), Citrix Systems (XenDesktop), Quest Software (QSFT) (vWorkspace), and Microsoft (VDI Suite), plus a handful of private vendors.
Microsoft, with its traditional strength in the desktop, partners now with Citrix Systems and Quest Software. VMware goes it alone by leveraging its strength in server virtualization.
Ultimately, this is a battle Microsoft can't afford to lose, given its focus on all-things-desktop, and its very limited success with server virtualization (Hyper-V).
I'll publish commentary on the strength/weakness of each vendor's offering in a future post, but for the purpose of this market commentary, one can assume these near-term use case wins are going to be a knife fight.
In summary, today VDI is a much more expensive version of Terminal Services. Because of the lack of TS application compatibility (i.e., not every application can be delivered via TS), VDI is being used to deliver a more robust set of apps to the same audience.
But make no mistake, VDI isn't cheaper than the current technologies and/or processes that today serve most of those 300 million desktops, and therefore, is unlikely to come near matching its prevailing hype/expectations.
Over time this will change -- desktop virtualization technology will mature and add features that make an end user want to use it (like off-line support and dynamic application provisioning).
However, we expect VDI to go from the top of the hype cycle where it is today, to what Gartner calls the trough of disillusionment, before becoming a truly big market.
Copyright 2011 Minyanville Media, Inc. All Rights Reserved.
Daily Recap Newsletter