Desktop Virtualization on Ulitzer

At the end of last year I made some predictions for how I think desktop virtualization will develop in 2010. People who have listened to Brian Madden TV's prediction show will have heard references to some of them and I think the time is right to share the whole list with our broader readership. I am interested in your thoughts too, feel free to comment whether you agree or disagree.

1. The proof year for hosted virtual desktops
From talking to our numerous customers and partners around the world, we see there is now general acceptance that the componentized desktop model is the way to go, and there is a good understanding of the sorts of benefits we will see.  We believe 2010 will see a distinct move to non-persistent, component-based desktops in large enterprises as challenges such as how to manage the user experience in a standardized desktop environment are solved by software providers.  We see more and more enterprises moving from PoC to full implementation.  As they say ‘the proof of the pudding is in the eating' and 2010 will be the proof year for hosted virtual desktops.

2. DV implementations will move from PoC to full implementations (the first >20K seat implementation will occur in 2010)
With over 4000 customers around the world, we are seeing many of our PoC implementations start to convert to full roll outs.  This is important because it means the business case for virtualized client computing is being proven by delivering independent, quantifiable, benefits.  As with any new technology, it is only at this stage that it becomes easy for the majority of organizations to make decisions on future technologies.  We believe that the first 20,000+ seat implementation of a virtualized desktop will go live in 2010.

3. Client virtualization starts to take off
The principal area of innovation in 2010 will be client virtualization, specifically virtualization of laptops. Client hypervisors from Citrix and VMware will ship and join those from Neocleus and Virtual Computer. These products support very different management models and the big debate of the year is going to be ‘how do we actually want to manage users on virtualized clients?'.  The hypervisor itself is only a small part of a client virtualization solution and most of the benefits come from changes to how we manage the platform rather than the introduction of the hypervisor itself. The model that will be ultimately successful will be based on componentization, similar in many ways to the model that is being implemented now in hosted desktop virtualization, but with some modifications. The modifications will be necessary to take account of the basic difference between a mobile device such as a laptop and a server in the data center - intermittent network connectivity. Key to success will be to preserve the essential features of componentization such as getting economies of scale across software components and delivery of the user environment in this more challenging platform. There are several techniques that are good candidates for this and we will see active debate throughout the year on their pros and cons. Given the proportion of laptops in business today and the imperative to get to a single management methodology across both hosted and roaming users, this will be an important and lively discussion. One of the outcomes will be the importance of choosing a user environment management solution that delivers across all of the platforms in use in the business be they hosted or client, virtualized or traditional.

4. The first commercially viable UIA solution
Since we believe that the non-persistent, component-based desktop model will be the prevalent virtual desktop model in 2010, we also believe that the emergence of a commercially viable solution to manage users ‘personal' applications will happen in 2010.  By standardizing the corporate desktop and separating it into its constituent parts, (corporate OS, corporate apps and user), the problem with non-corporate user apps remains.  User-installed applications in a non-persistent desktop model is a real challenge and cannot be solved easily.  The inevitable adoption of the non-persistent desktop model will bring a solution to market in 2010.

5. DaaS will begin the move to a more economical non-persistent, standardized desktop model
DaaS solutions out there today are based on a 1:1 user to desktop model.  Although many organizations benefit from subscription-based desktop management, the costs associated with managing and storing many unique desktop images will inhibit the adoption of this model.  By moving to a standardized desktop and separate user personality, significant economies of scale will be realized and DaaS will become a reality for more organizations.

6. Cloud Computing will show little progress in client computing (DaaS being just the first step and the only area of activity)
The Cloud Computing model for client computing will take several years to progress to a mainstream solution, depending primarily on Cloud Orchestration standards and high availability bandwidth.  We do believe, however, that the advent of non-persistent desktops in the newly-emerging DaaS model will begin the move of Cloud from theoretical hype to serious business case consideration. However DaaS is only part of the Cloud model of client computing and there is still plenty that needs doing to enable integration of services across multiple service providers

7. BYOPC morphs into HAPC
While the concept of BYOPC is attractive, in reality, this model is unlikely to be widely adopted by employees.  Support and warranty issues will cause unnecessary headaches for users and many will inevitably contact their own company support desk for assistance anyway.  What is more likely is the use of desktop virtualization in HAPC (Home Access PC).  Employees will be more likely to use their static home PC for work use out of hours and keep their work PC at the office.  A non-persistent virtual desktop model works well in this scenario, as long as the employee has a predictable and personal experience across both devices. These users will be served by either hosted solutions(SBC/HVD) or local execution, the ‘new thing' here is local execution.

8. Desktop component delivery commoditizes
As non-persistent corporate desktop images become the norm, the standardization and automation of the delivery of these assets will become less valuable and the updating of the source image and the management of the user's individual environment will increase in value.  Platform providers will deliver efficient means of delivering desktops and corporate applications from a standardized source image, with the management of these source images becoming the value-add in the mid-term.

Longer term, image management will move to become a service that is provided by software vendors and will be part of our contracts with the vendors. Configuration and personalization of applications will remain inside the business because so much of the information involved is business and user specific. This becomes a major direction for user environment management.

9. We reach a decision point on the mobile platform for client computing - virtualization or a de-facto standard. - Updated
Currently there are few corporate applications available to run on phone platforms. In part this is a result of the platform variance that has existed to date which would have required an application developer to support each platform separately. For a number of years vendors have been working on hypervisors for mobile devices that would allow a single common view across all the devices. VMware bought Trango a year or so back and investment was rife. This year we will either see virtualization get adopted for mobile platforms or a single platform adopted instead. If not virtualization what would the platform be? Android is too new, Microsoft suffers from platform variance perhaps leaving Apple or RIM to become the de facto standard for mobile devices - there is a sizable chance of Apple ousting Microsoft because of a lack of consistency in the Microsoft platform. Whoever moves to take this position will have to find a better, more business focused, way of selling applications than AppStores.

10.  Stand-alone application virtualization ISVs will find 2010 a real business challenge
Application virtualization is an increasing important component in desktop virtualization, and this will continue through 2010 and beyond.  Such is the importance of this mechanism, that we expect smaller, independent ISVs specializing in this area to become either swallowed up by the platform vendors (like Softricity and Thinstall) or find new directions.