In Why the Hottest New Tech Solution Will Go Cold
, I covered a brief overview of virtual desktop infrastructure (VDI) -- and in so doing, hoped to draw attention to the associated hype compared to market reality. In summary, I believe that while VDI will eventually become a big market, today the technology is materially over-appreciated and overvalued.
Simply put: The VDI hype far exceeds current market reality, and has for at least the past 12 months.
I will now outline the broader desktop virtualization (for short, DV) market, itself a superset of VDI, and provide an overview of each primary vendor’s DV strategy. Five years from now, my firm, Marker Advisors, believes some form of DV will touch almost every business desktop/laptop/netbook, making the DV market of far greater value than its subset, VDI.
But back to today’s reality.
Currently, there are four public companies chasing the DV opportunity: Citrix Systems
(MSFT), Quest Software
(QSFT), and VMware
(VMW). DV encompasses three desktop delivery methods (it’s really three different, related technologies, only one of which is VDI):1. Terminal Services (referred to as TS):
server-based, shared applications that send graphics to a variety of computing platforms. This technology is very mature and exampled by Microsoft’s Terminal Services (now called Remote Desktop Services) and Citrix Systems’ XenApp.
Advantages include its low cost (a large number of users can hit a single server), reliability, mature delivery of information, and a large partner/customer base.
Key disadvantages are limited application compatibility (delivers mostly Windows applications) and a lack of end-user configuration. There must also be an operating system on the client.
It's best used as a secure remote substitute for a user’s standard desktop, although it has also been employed as a low-cost desktop for task workers.2. Application Streaming (AS):
This technology streams (downloads) an application(s) to a user’s desktop on demand, and then runs it locally. An image is kept on the server (so it’s easy to maintain), while a copy or partial copy is sent to the desktop where it runs on the local operating system.
The primary advantage (and it’s a big one) is that applications can take advantage of the local computing power, yet still be managed by IT like a server-based application. It's also secure by nature, as the streamed application is isolated from the other applications on the desktop. Examples are Citrix Systems' XenApp (includes application streaming as part of the TS license), Microsoft App-V/MED-V, and VMware ThinApp.
A main disadvantage is longer initial application start-up and shutdown times (application streaming/syncing time), and that an operating system must be present at the client for the application to run. 3. VDI:
This technology is also a server-based computing solution, however, it remotes graphics for both the application(s) and the operating system. It differs from TS in that each user actually has their own desktop delivered from the server (TS provides shared applications). This means one doesn’t need a client operating system, or if they do, VDI will run as a “desktop within a desktop”. The advantages (secure, fast provisioning, multi-computer access) and disadvantages (very high cost, lack of end-user benefits) are more thoroughly detailed in Why the Hottest New Tech Solution Will Go Cold
.Vendor capabilities (in alphabetical order):
Vendor strategy, from cruising altitude:Citrix Systems
- Citrix Systems offers all three -- TS (XenApp), AS (XenApp), VDI (XenDesktop). Citrix Systems, in its enterprise software business, is as close to a pure play DV vendor as exists today.
- Microsoft offers all three -- TS (Windows Server Remote Desktop Services), AS (App-V, MED-V), and VDI (VDI Suite). But one should note that at the present time, native Microsoft capability in TS and VDI scale poorly. As a result, Microsoft partners with Citrix Systems or Quest Software to be enterprise-capable. Microsoft has a huge desktop business, and is playing catch-up in this technology shift.
- Quest Software offers TS and VDI -- The company offers an enhancement to Microsoft TS and offers VDI (vWorkspace) on any platform. Quest Software has a very diverse line of enterprise software management tools, and vWorkspace is a small but important division.
- VMware offers AS and VDI -- VMware offers AS (ThinApp) and VDI (View), but no TS offering. VMware dominates the server-virtualization market with vSphere, however, some pundits argue the company “just doesn’t get the desktop."
Citrix Systems has a complete, integrated stack of technology for delivering DV (TS, AS, VDI) -- including data center, network, and desktop. In my opinion, Citrix Systems offers the most complete end-to-end solution on the market.
XenApp revenue (TS, AS) has been hurt by the recession, in concert with the decline in the number of remote knowledge workers. To combat this erosion, Citrix Systems has positioned its VDI offering XenDesktop at the top of its product line (the high-end license includes TS, AS, and VDI). This represents a clear strategy designed to upgrade the vendor’s huge XenApp (TS) customer base to its full DV suite.
To simplify (perhaps oversimplify) Citrix Systems’ strategy, the company is planning to upgrade a material chunk of its large customer base from XenApp licenses to XenDesktop licenses, at a higher average sales price. Although XenDesktop can run on theXenServer (Citrix Systems), Hyper-V (Microsoft), or vSphere (VMware) hypervisors, Citrix Systems would prefer to sell its entire stack of technologies for DV, from the data center to the desktop. Were this any other ISV, it would be a solid strategy.
However, Citrix Systems brings a whole new definition to the term “co-dependent”.
This strategy ultimately puts Citrix Systems’ partnership with Microsoft at risk. When Citrix Systems sells its full product line, it competes with Microsoft rather than complementing it. (A future article about desktop virtualization will elaborate on the fine line Citrix Systems must navigate with Microsoft.) VMware
VMware nearly has the complete DV stack (it's missing TS). The company has a pervasive server farm platform (vSphere), and View is a capable player in the VDI race. VMware has recently started pushing AS (ThinApp), as interest in this form of virtualization is accelerating.
VMware’s strategy is to sell View to the vSphere faithful, of which there are many. In my opinion, VMware privately sees View as just another mobile workload -- another way to sell vSphere server software. In this respect, the company is partially correct.
However, VMware is the poster child of what I believe remains the major impediment to the eventual success of the DV market: the lack of end-user benefit messaging. VMware still perceives this as data center technology, and pushes an IT-centric sales message.
Ultimately,VMware’s vision is to put its desktop hypervisor between the hardware and the operating system, in very much the same way it has on the server. That vision may be some years away, if ever; though I imagine it’s never far from management’s dreams. Or lips.Microsoft
Like Citrix Systems, Microsoft also has a full stack of DV technology. However, Microsoft has trouble scaling -- both in terms of number of users and over slower connections.
As one would expect, Microsoft can outfit a DV server farm on Hyper-V. Yet to solve its scalability problems, the company has been handing VDI leads of 100+ seats (or those using slow connections) to Citrix Systems for the past year or so. Microsoft is clearly comfortable with this, as it has been giving Citrix Systems leads for large TS installations for about 10 years.
However, that exclusive VDI relationship has come to an end. Quest Software has recently joined the Microsoft lead party.
My opinion is pretty straightforward -- Microsoft must field a compelling entry in DV, and do so very soon. Its current product line is materially behind both VMware’s and Citrix Systems’, and it has a relatively short window (no pun intended) of opportunity to gain mindshare in this race. Its new partnership with Quest Software will help, but Microsoft must get serious about pushing its vision of DV, before VMware gains any more advantage.Quest Software
Quest Software has a very strong, hypervisor agnostic VDI offering; it runs on Hyper-V, vSphere, and others. However, the company has neither a stand-alone TS or AS offering.
What Quest Software does have (notably) is it removes the scalability problems of Microsoft’s VDI and TS offerings. In other words, to paraphrase Jerry Maguire, “[Quest Software] completes [Microsoft’s DV suite].” When partnered with Microsoft, it's easiest to think of vWorkspace as an enterprise add-on for Microsoft’s TSand VDI Suite.
In my opinion, the addition of vWorkspace on top of the Microsoft VDI Suite gives that partnership almost as full and scalable a solution as Citrix Systems, and easily on par with VMware. It at least gets Microsoft into the DV major league. Bottom Line
As I outlined in Why the Hottest New Tech Solution Will Go Cold
, I believe that in its current iteration, VDI is over-hyped and overestimated.
Investors should really be asking the question, “Are there any material net new users of VDI? Or is it just a more expensive technology that’s upgrading the TS user base?”
This is particularly relevant for Citrix Systems, for if the VDI market today is mostly upgrading the TS user base, then as Citrix Systems loses a VDI opportunity, it could also be losing XenApp (Terminal Services) seats -- the lion’s share of its software revenue base.
On the other hand, I believe that application virtualization (AS) is under-hyped and underestimated, and moreover, has a material opportunity to add net new users as part of the Win 7 migration. Our forecast from Marker Advisors is that most desktops/laptops/netbooks will run some form of virtualization, but the majority will use TS or AS on top of an OEM-provided operating system.
At the end of the day, we don’t see the OEM model of O/S distribution being replaced by virtualization, but rather augmented by DV technologies.
As this market plays out, we believe Microsoft, with the help of its partners Citrix Systems and Quest Software, will gain material share. VMware will win in enterprise accounts that are trying to be true to the vSphere platform. However, in the end I expect VMware to have only a small fraction of the overall DV market (if perhaps a healthy percentage of the smaller VDI market).
Citrix Systems is the pure play of the group, and the one that must at least “place” in this race. Either the vendor succeeds in DV and grows its user base beyond XenApp (that is, it needs net new users, not just XenApp users that traded up to XenDesktop), or it will have a faltering enterprise software business.
In this regard, I've heard many Citrix Systems partners characterize the company’s assault on DV thus, “[Citrix Systems] has burned its ships on the shores of DV.” You have to love that.
However, while we will likely soon see a flurry of renewed energy around the Citrix Systems/Microsoft partnership, inevitably, we see Citrix Systems having to compete with everyone. Yes, including Microsoft. Why?
Because to win, and meet the expectations of the Street, Citrix Systems will have to push its entire integrated suite in competitive deals, which will eventually drive a wedge in its Microsoft VDI partnership. It’s already begun. Adding to this complexity is the fact that there is significant overlap in the two vendor’s partner/reseller channels; for most of these partners, we’ll let you figure out which side of the bread has the butter.
I will elaborate on the interesting (and developing) dynamic of the Microsoft/Citrix Systems/Quest Software partnership/love triangle in a future Minyanville article.