VMware’s Desktop Strategy – Doubling Down on VDI

Our analysts leave VMworld 2010 having had great fun, having met some most excellent people and having been impressed with interesting vendors. Yet one thing puzzles after attending and discussing what we’ve experienced.

What is the focus of VMware’s Desktop Strategy?

  1. Is VMware really committed the Desktop Virtualization Market?
  2. What is VMware’s strategy going forward?
  3. How will VMware Differentiate from the Competition?
  4. How will VMware compete with the new vendors looking to disrupt and reinvent the desktop space?

Why do we ask these questions?

2010 has so far been a year of mixed progress for VMware’s Desktop initiatives. The good news is that more of the pilots that have been started in the last few years are being deployed into production. VMware also shipped View 4.5 a major feature of which was to remove the “experimental” label from the “Off-line Mode” features. However there are also been setbacks and concerns.

There have been changes in the leadership of the Desktop Virtualization Group which brought into question VMware’s strategy and even the intensity of VMware’s commitment to the desktop delivery space. Virtual Profiles (acquired as a part of the RTO deal) was originally going to be part of View 4.5, but then got pulled from the beta as the product went into production. This had the effect of highlighting that VMware understood the importance of the problem of user settings management in centralised environment to View’s users, but then yanked the proposed solution out from underneath their feet. There is still no integration of PCoIP with Security Server which means the options for delivering remote access to users is reduced, and there is no native support for optimising the remote protocol for high latency connections. Finally, there were rumors that VMware had “killed” their bare metal client hypervisor solution, or Client Virtualization Platform (CVP): rumors that were effectively confirmed at VMworld (as we anticipated) there are no plans to productize CVP at this time.

VMware announced it is doubling down on VDI – but is that the best place for VMware to go? Is it a whole lot of puff, without an amount of huff that is unlikely to blow the competition down?

Is VMware Really Committed to the Desktop Virtualization Market?

VMware has stated it is committed to its approach to the market, but that is not the same thing as being committed to the market as whole. VMware is clearly continuing to invest in its core VDI offering, and has now released Off-line Mode as an officially supported method to handle the use case of the occasionally disconnected user. In addition, VMWorld saw the announcement that VMware considers  View 4.5 Premier Enterprise Ready. When VMware View 4.0 was assessed against Burton Group Server Hosted Virtual Desktop evaluation criteria there were a number of significant shortcomings that were deal-breakers for many large enterprises:

  • Role-based access  controls (RBACs) for delegation of administrative duties
  • Administrative change logging capabilities to provide an audit trail for all administrative actions
  • Official support for Windows 7 operating systems
  • Enterprise management software integration

These issues have been addressed. There is new management at the top of the Desktop unit, and there are acknowledgements that improved management and more is coming. VMware is making the case that it should be the enterprise desktop and application delivery platform of the future.

What is VMware’s Strategy Going Forward?

The key thing to understand about VMware’s strategy is that it is very VDI focused. If anything, VMWorld indicated a doubling down on the part of VMware on VDI. So … even more focused on … focused? In what way? The strategy sounded exactly like the one that Citrix articulated in 2005 which was “any application, on any device, at any time”.

In 2005 Citrix wanted to put ICA clients on any device that was a credible recipient of a Windows application. Why? because for you as a business it helped justify your investment in their product. It wasn’t just about delivering applications to desktops, but the concept of delivering data to users wherever they were, regardless of the device.  An investment in your desktop delivery not only saves on the desktop delivery – but expands your delivery options. You were investing not only to reduce costs now, but to save on costs in the future. What a great concept.

This is exactly what VMware intends to do with the View client. Sure – its a great concept. VMware made specific statements that it intended to be the “glue in the middle” that allowed applications from Microsoft, Google and Apple to run seamlessly on devices from Microsoft, Google and Apple. But, this amounts to nothing more than a strategy of putting the View client on multiple platforms. This sounds like VMware’s strategy is… do what Citrix does now?

You win new markets through marketing and innovation. Is doing what Citrix did is innovating?

At the moment, VMware is not engaging in some of the extended versions of the Desktop Virtualization Market. In particular VMware has walked away from a client side hypervisor (shelving the plans to productize CVP) leaving these initiatives to Citrix and Virtual Computer (and possibly Intel with the rumored acquisition of Neocleus). Specifics aside, VMware is not being nearly as aggressive as they could in the area of using desktop virtualization as a catalyst to reinvent how desktop computing is managed. Broadening access to desktop applications across platforms is interesting, but it does not have nearly the organization impact and ROI associated with it as would re-invention of the desktop management paradigm. If you build it, they may well come – but the trick is to make them stay there and come again. VMware clearly intends to reinvent how servers and server based applications are managed, and one wonders why it does not have the vision or the stomach to attempt the same for the desktop – leaving changing the paradigm of desktop management to the likes of MokaFive, Ringcube,Virtual Computer, Unidesk, and Wanova.

How will VMware Differentiate their Desktop Offering from the Competition?

VDI does not address any significant use cases that have not been addressed by Presentation Virtualisation solutions. You could argue VMware coined the term “VDI” because they needed a product that didn’t fit into the (then) standard Server Based Computing environment which we call Presentation Virtualization (PV). With VDI there is a difference to PV, in that opposed to having N copies of user applications running together in one back end server OS, there is one desktop OS running in the data center for each user. This avoids compatibility issues with desktop applications and a Windows Server OS, and can reduce issues with desktop applications being incompatible with each other. Yet otherwise both represent the idea of running desktops in the back-end data center and connecting to these desktops with remote access protocols like RDP, PCOIP, and ICA.

However Citrix, for example, has a suite of solutions with XenApp, XenDesktop, and the soon to be released XenClient, being options (along with server and desktop streaming, user virtualisation and applications virtualization). Citrix can offer an organisation the option to manage its desktop environment. As we’ve mentioned Citrix has long played the heterogeneous device game with support for Windows Desktops, a very broad range of thin-clients, and smartphones including Apple iPhone’s and iPads, Google Android devices and RIM’s Blackberry. The only issues for Citrix are that its desktop service solutions are not integrated – while competitors, such as Ericom, Quest and 2x offer simpler, cleaner, and more easily administered solutions which are more importantly cheaper to license. Customers respond to cheaper. VMware is being very aggressive on the pricing front, whereas Citrix is still trying to charge premium pricing for its premium XenApp/XenDesktop products. But, VMWare only offer VDI.

How will VMware compete with the new vendors looking to disrupt and reinvent the desktop space: is VDI a Yesterday Technology?

Consider the desktops in use in your organization, and look at the applications they access and how they access them. It is more likely that you have a Windows fat-client application based environment. You may use thin-clients, but it is likely those thin-clients support “fat” applications, and that you need to support other devices as well. Even if we project forward that there will be a proliferation of devices like the Apple iPad and Google Android phones, there is going to still be a demand for traditional PC devices (either desktop or laptop) to deliver applications. That delivery mechanism may be VDI, it may be web-based, streamed applications,  it may be a mix – but in the majority of cases there will still be a need for enterprises to manage end-devices. It is interesting to note that existing VDI solutions often rely on the end-device having an up-to-date Windows installation in order to deliver the best remote experience for VDI. Zero clients are available – but like “coke zero” has a tiny amount of calories, so a zero client has some need for management.

New vendors in the desktop space are not solely focused upon improving upon Presentation Virutalization or Desktop Virtualization, but rather they are looking to change how desktop computing is managed in the first place. To understand the opportunity one must first understand the problem. The problem is that IT wants (and needs if you want to transcend to a Cloud service for desktop or application management) to manage, and be able to understand the cost,of supporting desktops.  To uniformly manage and support is one thing,  but to allow users to customize their desktops and laptops with settings, applications and data that make the device useful to them is another.

These two desires are fundamentally in conflict, and addressing this conflict is the biggest opportunity in desktop computing.

Let’s take a brief look at what some of the really interesting vendors in this space offer towards addressed this challenge:

  • AppSense, RES Software and TriCerat offer solutions to manage the user environment regardless of the operating system and to persist those settings between environments. These vendors allow you to have users move between traditional desktops and virtual desktops as seamlessly as possible. In addition, RES Software have recently announced their reverse seamless technology. AppSense are developing User Application Installation services. Both companies work across desktop delivery environments – be they presentation or desktop virtualization or traditional desktops.
  • Symantec offers its End Point Virtualization Suite which combines application streaming and applications virtualization features with support for desktop deployment options including VMware, Citrix, and fat client desktops.
  • Virtual Computer has carved out a clear early market leadership position in the area of using a client side hypervisor as the platform for a management suite distinguishing the myth that client hypervisors are only useful as a “Bring Your Own Computer” (BYOC) solution. Virtual Computer’s NxTop is coming up to its third major release. Citrix are about to release XenClient – you may  presume this would be the most feature rich solution on the market. Our comparison of initial releases from Citrix and VirutalComputer showed this was not the case. The distinguishing features of NxTop are that it is a client hypervisor with broad hardware support, and it also includes the management software to allow a service to centrally manage, backup, provision, and update desktops and laptops.
  • MokaFive and RingCube both take the approach of layering a centrally managed desktop on top of what the user already has. This has the advantage of not requiring a device with a client side hypervisor installed as the starting point.
  • UniDesk and Wanova  separate out the operating system from the applications and the user environment and then allow each to be managed separately. While UniDesk only currently support a VMware VDI environment, Wanova offer a solution that can be used across the desktop estate while UniDesk are in a better position to help VMware View adopters manage their environment more effectively.

If I wanted to Go There, Would I Start From Here?

VMware has said that is it committed to its Desktop Virtualization Strategy but VMware’s commitment to VDI as the only solution is going to mean that unless you are only going to deploy VDI, you’ll likely consider another vendor to help you achieve your desktop delivery goal. A VDI only desktop strategy for a business is not common – but it does not fit many markets, it does not offer a solution that can be used in to all organisations.

If VMware’s desktop strategy going forward is to focus on VDI, to double down on VDI, is that good? Considering the limitations of such a strategy for those designing their desktop solutions, is is a risky strategy. Why is it VMware’s strategy? If the hypervisor is a going to be a commodity (you don’t pay for the hypervisor, you pay for the management of the hypervisor), is offering a desktop management service a market VMware wants to be in – given they’ve a great deal of catch-up because they only support VDI as a solution.

Organisations have two conflicting options going forward, and addressing this conflict is the biggest opportunity in desktop computing. Can you manage the demand for users to have effective IT at their fingertips while controlling access and costs from the centre?

A strategy can involve a number of options; attack if you think you can be stronger and win outright; to bide time to regroup or assess the situation; and indeed, to retreat or surrender are valid strategy options as well. VMware had a great initiative with VDI which has in turn caused a massive upheaval in the offering of products available for desktop services. VMware don’t appear to have noticed this massive upheaval yet.