What are the real-world non-BS reasons people use VDI today?

Yesterday I had a discussion with some business-type people who weren't terribly technical and who were interested in desktop virtualization as it relates to power savings. They were generally familiar with server virtualization and equated desktop virtualization to server virtualization plus thin clients, which I guess is VDI.

Yesterday I had a discussion with some business-type people who weren't terribly technical and who were interested in desktop virtualization as it relates to power savings. They were generally familiar with server virtualization and equated desktop virtualization to server virtualization plus thin clients, which I guess is VDI. So anyway, they asked me point blank, "Why do people use VDI today?"

My first instinct was to go into my whole elevator pitch about the advantages of server-based computing in general (Management, Access, Performance, Security) and then the specific benefits that VDI has over Terminal Server (same OS as desktops, more granular resource management, better app compat)... But before I could form my words, I had a realization that these folks didn't really care about about all these academic reasons—they simply wanted to know why real companies were using VDI today. (Not that the academic reasons aren't true.)

I shared the following three real-world reasons with them:

Same benefits of Terminal Server / SBC, without the hassle

I've made the argument before that most of today's VDI environments could probably run fine on Terminal Server, but that's not the point. (Remember, our point today is to talk about why people are using VDI, and there are certainly a lot of people out there who want the remote computing advantages without the hassle of Terminal Server.) And sure, we could talk about Terminal Server having a better user density, but the reality today is that a lot of people don't care. To them it's not worth learning a whole new OS (Terminal Server) just to save on some density.

So this means that we've got people using VDI for regulatory compliance, classrooms, and ease of image management. And the reality today is that these people want datacenter hosted desktops without the "hassle" of Terminal Server, and for them, VDI is the answer.

Virtualization spillover from servers

I can write all I want about how I wish that today's VDI deployments were being designed by the old-school desktop people instead of server virtualization folks who are looking for more things to virtualize. The the reality today—like it or not—is that some people are using VDI simply because it has "V" in the name. In other words, they like the benefits they got with server virtualization and they want to extend that to desktops, irrespective of the fact that desktop virtualization and server virtualization are not the same thing.

Flexibility for DR / Pandemic

I guess this is another benefit that also applies to TS-based solutions, but I know for sure of several companies that have built out VDI environments to enable employees to continue working in disaster scenarios. And some of these were even used a few months ago during that whole H1N1 scare where people were saying, "If you're coughing, don't come into work." VDI allowed those workers to connect from home.

What do you think?

Frankly there's a lot of hogwash in this article. Some of these reasons are misguided and/or things I don't agree with. But when looking at the real world today in January 2010, I think these reasons are among the most common as to why customers are using VDI. What do you think? Did I miss anything? What are the real reasons you or your customers are using VDI today?

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

In my organisation it comes down to several key points:

1. Session Portability - The ability for our staff to travel from one end of the campus to the other and pickup their existing session is a strong +

2. Thin Client Endpoints - We no longer need to send techs out onsite to look at issues, If it cant be done remotely then its going to be a simple swap out of the end point.

3. The chance to right a few wrongs and provide single sign on functionality to our end users.

4. Application Streaming - To make VDI really work you need to stream/virtualize your applications into your templates. This allows us to negate a lot of the application compatibilty complexities in our existing envrionment.

5. OS Flexibility - Following on from Point 4, the days of old style desktop rollouts are gone. Providing a user OS upgrade/downgrade is no longer a huge effort as in the past.



Jeroenk, I think your post validates what Brian is saying. All of your points are achieved with traditional SBC.

To be fair, I'd ignored CItrix for 7 years despite some great success with MF in the late 90's. We use VMWare heavily and this year I get very interested in VDI to address a number of new challenges. I engaged with VMWare and continued to ignore Citrix until I figured I may as well have a look. 4 months later, I've got "XenDesktop Platinum" licences plus a CAG and various repeater appliances.....all of which I'm going to use for XenApp, at most a couple of exceptions for support people who need ability to do whatever they want / be unmanaged.

My conclusion was that I also got sucked in by the 'V' word but realised that SBC remains more than relevant, easier to manage and much more cost effective for my needs. I suspect my needs are fairly typical.


Security and speed.

To be honest, i'm based in Minneapolis and I've been using a hosted virtual desktop located in Ft. Lauderdale.  And here are my benefits:

1. Security: My data is now in the data center.  I don't have to worry as much about having my laptop messed up and lossing all of my data. When i travel, i sync certain items down (namely the docs I'm working on). Bu as my battery only lasts about 1 hour, I don't expect to do much work on the plane anyways.

2. Speed: I have a fast broadband connection, but that still doesn't really impact the speed when I'm working with systems in the data center.  I've noticed that the applications run much faster as the backend data is now in close proximity to my virtual desktop (no surprise here).  It is a nice change.

Is my virtual desktop perfect? Not yet. There are some items I wish worked differently, but given time, I'm positive they will be improved. Details to come



You were close with "Same benefits of Terminal Server / SBC, without the hassle" but I think you did not hammer the nail on the head.

VDI makes sense in distributed organizations that want to centralize and consolidate, and have a relatively many applications in comparison to the total amount of users.

Often these organizations have many apps which are only used by a very small number of users. The more diverse the application set and the more diverse the requirements, the more interesting the business case for VDI becomes. In shared environments like Terminal Services, applications with only a few users become expensive and complicated. In a VDI world the impact of such apps is dramatically lower. I think this is the essence of “Same benefits of Terminal Server / SBC, without the hassle". So, I agree with you: but I would formulate this differently.

The thing is, VDI is not the only ultimate solution. There are many valid reasons to prefer SBC, and many reasons to prefer a classic PC/Laptop infrastructure. Honestly, the magic truly happens if you combine the different desktop models. Each desktop model have their own unique advantages, and combining these models allow you to broadly support desktop requirements.

PS: @natd, I completely agree with you, this is also valid for SBC.


I don't understand these claims that you need App Virtualization to make VDI work.  Not only in this post but in some other topics as well.  If you're already using something manage physical desktops (SMS, Altiris, SCCM, etc) and VDI has no app compat problems compared to TS they why the heck do I need to throw another layer of complexity and a resource who can sequence, package and maintain virtual apps???  I completely understand AppVirt in the TS world, but not so much in the physical and VDI world.  Do your users really need to be able to run Office 2000, 2003 and 2007 at the same time? (sorry I use that reference because its the same sales pitch every appvirt vendor demos when they show their product)  If anything in my experiences with AppVirt, you now add a small level of app compat issues with application that need to interact with one another (like MS OCS and Outlook, IE and Adobe, etc)



"Do your users really need to be able to run Office 2000, 2003 and 2007 at the same time?"

Yes this is where App virtualization is a major plus. We are a large University with strong links and partnerships with other institutions and businesses all running various versions of the same apps (not just office) so we use virtual apps get us around compatibility issues, It's not just our partnerships that we take the hit on. We're separated into several faculties with many departments all run as separate "business units". So these issues are internal as well.

Please remember developers, admins and web devs. Running several versions of the same app (browser, etc) speeds up development time 10 fold.

App virtualization in the physical world is key in providing a flexible teaching environment (specially where dedicate lab space is slowly vanishing). This can't be achieved with traditional solutions such as SMS, ZenWorks, etc. And for many of our Apps TS isn't good enough. VDI is starting to bridge the gap is this space.

The whole app chaining issue is starting to vanish with virtual apps so this isn't really an issue anymore.

Virtual Apps complement virtual desktops the same way butter complements bread. Now you can dynamically assemble a desktop based on what the user requirements. Rather than providing a legacy bloated desktop in a virtual world.

We're using VDI as a catalyst to "rethink" the desktop. Not bringing legacy solutions forward.

It sounds like you were stung with early versions of Softgrid - like we were :)


You can pound technical reasons into the sand until you turn blue in the face but the real driver is how much am I going to save. Most companies we work with are sick of the desktop management costs and efforts associated with yesturdays desktop computing. VDI brings many technical advantages as you've explained, but knowing presise cost to implement VDI to the right groups of people as well as amounts of FTE's they could cut or repurpose is really the driver most companies are looking for.



I agree that for 99% of Businesses out there, TOC and ROI is a big deal...

However in many Universities this is a bonus driver. The real deal is about adding value to the end user.

Most solution (in UK education anyway) purchases very rarly take TOC into account.

Having said that we in the UK edu sector will be taking big budget cuts in the next few years. So my stance on this may change.

Anyway this is off topic so I shall shutup (for now).



I didn't say there was no value at all to App Virt, you obviously have a use case for it.

My problem is in saying VDI doesn't work right without it.  

I'm not in the education market, but I've consulted in the past for a couple schools on the East coast and I can see how what you're doing makes sense.  However in the chemical manufacturing industry I'm in now, there is little to no value.  Most of our apps are web based and the only fat client upgrade process we tend to go thru is Office.  For the most part all our developers use green-screens for AS/400 work or are Sharepoint/Visual Studio developers

This tends to fall in line to what @Ryan just said.  I can definitely save money with SBC, I can kinda sorta justify saving some "soft dollars" with VDI.  But I can't justify the FTE to manage virtual apps.  I'm not taking anything away from the education sector and I've met many brilliant IT people that work there, but there are some significant difference between that sector and the enterprise corporate sector.

With all that outta the way...my non-BS reasons for VDI...like @DanielFeller said - Speed and Agility.  We have a support staff in India that primarily supports systems in the US, having a VDI in the same datacenter as what their supporting speeds things up greatly.  Again being in the manufacturing industry we have segregated process control networks and systems that monitor reactors, dryers, etc and using VDI also fills some compliance gaps with outside vendors and contractors.


I just realized using the term "Same benefits of Terminal Server / SBC, without the hassle" is both true and utterly false.

The reality of VDI today is that from an end user and application point of view this statement might be true.

But still today, from an infrastructure point of view, VDI is complicated and in reality quite a hassle to make it work. This is similar to SBC implementation 5-10 years ago.  I dare to say that currently the VDI infrastructure challenges are far greater than the mature classic desktop and SBC implementations.



Right on. I see too many people looking at the desktop virtualziation debate from 1 side: IT.  There are 3 sides: Business, Users, IT.  

I'm not very fond of going down the ROI/TCO routes because so much of it is soft costs that are really hard to quantify.  I prefer to look at those items that can really impact the business:

1. Security

2. Agility

3. Speed

4. Management

With SBC based apps, it worked wonders if i was using a single app, but when i was trying to use multiple apps, it just didn't feel right. Just didn't feel like a desktop environment. Doing the hosted VM model, it feels different. It feels like a desktop, which is what I'm used to,which means I find that I don't complain as much :)

Twitter: @djfeller


One solid non-BS reason is organizational and operational alignment. With SBC customers have to change the way they do things. Many IT Organizations have clearly defined boundaries around desktop, application and server ownership. With SBC it’s a big mash-up of server, desktop and applications which forces organizations into restructuring and inevitably ends up in finger pointing when something falls through the cracks. VDI adoption doesn’t come with the same organizational angst which is a distinctive plus.



There's no one size fits all and never was.


@Tony (again :) )

It's true everyone’s use case is different. I can see the argument that 100% of apps don't need to be virtualized. Sometimes it can cut your nose off to spite your face.

Your lucky... We have just under 700 applications and that doesn’t include the web apps. Virtualization is pretty much the only model we can use without the risk of cardiac failure. I agree there are significant differences between corporate and education sectors. However you will often (but not always) find the most complex systems in education. I suspect that’s why MS, Novell, etc use us guinea pigs - We take their technologies to the edge.

I can't speak for non-uk universities but our students now pay substantial fee's. We like the corporate world HAVE TO deliver 99.999% uptime and have DR and everything else that goes with it. VDI has a massive part to play in this space also.

@Daniel the 2nd/1st (depending how you look at it)

I agree.


Hi all,

As much as I’m all over VDI (or SBC ) as such I see the problems; I keep repeating as a Duracell that all of us commenting here fully owns and controls their PC and working environment, and as such are never really capable of assessing the reality of those affected of the , willingly or unwillingly,  proponents.

Sure, there are use cases, even beneficial (all tech issues aside), but today’s reality is not quite so, today’s offerings deliver so and so. Nothing anyone of you would cope with. The others, right? The users, right? Not YOU, right?

For being in the business since way back, I never envisioned anything but remoting the burdensome applications, having my data secured, abstracted away from whatever my computer might be. In being a user, I gladly remote away those apps that have no rational reason to be at me. For my productivity, for my creativity – I am, and I embrace and utilize sooner than you say Hello!

This is my critique.


"Nothing anyone of you would cope with. The others, right? The users, right? Not YOU, right?"

I cope. It largely depends on how engage with the technology.

I see where your coming from. But depends on the application the user is using. If your doing excel and outlook then that type of "task worker" couldn't give a flying monkeys. They will pretty much get a perfect experience. - Though TS could be used here rather than dedicated virtual desktops.

I myself use VDI for 90 - 95% of my work. Anything else (multimedia, etc) I do on the end point (which is a laptop). But I class myself as a "power user". Even now multimedia performance is rapidly improving.

1 -2 years ago it was a different story but the technology is catching up fast. I think the next 12 - 24 months are going to be very interesting for the end-user experience.

@Al Grandville

I agree - VDI seems to be getting management thinking about technology again. I think its because they see it as an excuse to get rid of their bloated desktop and get new a laptop that they would otherwise find it hard to justify :)


I would think that is truly a factor of Money.

What  mean by that is hardware and software are cheap compared to head count.

If you have talented people who are good with and understand citrix or terminal server that would most likely be your solution.

However if your talent pool is within the Hypervisor be it ESX/v-sphere, Hyper-V, XEN, KVM, then your logical solution would be a VDI solution


Brian, Excellent post! I completely agree with your "same benefits as TS without the hassle", except, does VDI really do this?  Customers tell us traditional VDI brings a new set of hassles bigger than the ones it replaces: high upfront cost, unclear ROI, cultural change away from desktop IT, change in how they buy desktops (top-down as opposed to in phases).

What we hear from customers running TS is they want VDI as an upgrade if it can be at a price comparable to the cost of a PC, if it can be setup and run by desktop IT, and if it allows them to buy desktops in chunks of 25 to a few hundred.  This is the sweet spot we have designed for, and we have several customers that have upgraded from TS to VDI and are happy with the results.


We use VDI in a small business and it's not from the two supposed market leaders so often mentioned in these circles.

Ya know, if your a 'business type' and want generalized reasons why 'other' companies are currently using VDI, your not doing your due diligence.  There are so many use cases for all the numerous parts within a SBC or VDI solution specific to each business, these conversations rarely lend themselves to be helpful in a decision making process.  You have perform a 'deep dive' into the business internals (process, organizational, geographic, security, budget, etc...) and consider the 'academic reasons' to answer the question: why should WE use SBC or VDI.

VDI is a nascent technology compared to traditional SBC.  This means each vendor's VDI offering is likely to advance over time and offer new pieces to the puzzle of providing the best solution for your use case.


So here's the thing - everyone above is discussing VDI/HVD, not desktop virtualization as a whole.

Of course there are benefits of Hosted Virtual Desktops and for app virtualization, and of using them together, but that only covers a pretty narrow use case. How about streamed desktops with or without app virtualization? App virtualization on a fat/rich client? How about (soon to be available) type 1 client virtualization platforms?

As Frank Vandebergh writes above, no one size fits all. This is absolutely true - even within a single company. Take this scenario for example. Task workers on thin clients with XenApp based desktops and apps. Office based knowledge workers using streamed desktops and streamed applications. Remote (home/overseas) knowledge workers using Hosted Virtual Desktops, and of course road warriors on laptops that want to have full control of their machine despite the company's desire for a secure, managed environment. Then there is Disaster Recovery to think about...

XenDesktop covers all these (and more) scenarios under a single license and removes the vast majority the functional and end-user experience based objections to desktop virtualization - all that remains is the financial factor - TCO v Setup cost, Capex v Opex... Of course these are getting better all the time, sometimes it's just a matter of being able to see the big picture from a business point of view.