Is VDI driving a resurgence of thin client sales?

In the past I've written that VDI is "just a single user terminal server." But I've also written that one of the ways VDI projects fail is "thinking VDI is nothing more than a single user terminal server.

In the past I've written that VDI is "just a single user terminal server." But I've also written that one of the ways VDI projects fail is "thinking VDI is nothing more than a single user terminal server." Obviously these two statements are at odds with each other, so what gives?

To understand the discrepancy, we have to dig into the differences between Terminal Server/RDSH-based (Citrix XenApp, etc.) and VDI-based (Citrix XenDesktop, VMware View, etc.) projects. I'll use Citrix's product names in this article, but the concepts apply to any RDSH and VDI environments.

Because there are over 100 million Citrix XenApp users in the world, and the product has been around in one form or another for over fifteen years, many of us have confidence that "we know how to do HDX, we know how to do thin clients, etc." The knowledge we have is based on HDX and thin clients for XenApp. Most companies use XenApp either to (1) deliver a few corporate apps to users with existing desktops and laptops, or (2) deliver a complete desktop via a thin client to so-called "task workers" (call centers, shared terminals, etc.).

But when it comes to VDI, even though the HDX protocol and the concept of thin clients are the same as with XenApp, the use cases are wildly different. For example, users who only use XenApp for a few applications do their "heavy" work (web browsing, graphics, peripherals, etc.) via local applications running on their desktops. XenApp-based task workers with thin clients don't push XenApp and HDX to its limits.

But when it comes to VDI, it's more of an "all or nothing" thing for users. Companies don't use VDI to augment existing physical desktops—they use VDI to replace existing physical desktops. (After all, if you can't replace the local desktop, then adding VDI just means you're managing two desktops per user instead of one.) But the fact that VDI is desktop replacement this means that your HDX (or PCoIP, RDP, RemoteFX...) connection has to support everything that users want to do—web browsing, audio, video, peripherals, 1900x1080 resolutions, multiple displays, client side cameras and microphones, finger print scanners, etc.

Unfortunately companies quickly learn that the networks, protocols, and thin client devices that work fine for "just a little XenApp here and there" break down when companies try to use them with the power and capabilities required for day-in, day-out heavy desktop VDI usage.

"Everyone who needs a thin client is already using one"

When VDI first came on the market a few years ago, the thin client vendors started making devices that were specifically tuned to VDI use cases. I scoffed at these, writing that "everyone who needs a thin client already has one."

Yet despite my skepticism, thin client (or smart client, zero client, cloud client, whatever-they-call-it client) sales grew, and both Wyse and NComputing had their best quarters ever in 2012. (HP might have too—I just don't have that data.)

So again, what gives?

This is when I realized that the thin clients people were buying were for VDI and desktop replacement projects. For any existing XenApp projects, yeah, everyone bought that hardware five years ago and it still works fine. But the people who are using VDI have chosen it over XenApp because they have some specific advanced need—they're doing that desktop replacement or that advanced powerful application or something like that. And in their cases, the thin client device from five years ago just won't cut it. They need dual displays and USB2 and client side audio and the ability to plug in phones and cameras. (In a sense, they need a desktop!)

By the way, it's also interesting that the thin client devices sold for today's desktop remoting are not necessarily more expensive. Sure, they have more capabilities to satisfy the requirements of the full desktop VDI-based remoting, but in terms of cost, HP, Dell/Wyse, and NComputing all have HDX-capable devices starting under $200, and for $300 you can get something that supports multiple 1900x1080 displays.The bottom line is that people go to VDI because they want datacenter-based desktops to do things that RDSH and XenApp can't do, and that requires a different thin client than what they purchased in the past.  But fortunately those thin clients are available today, and they're selling like crazy!

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

Always amazed when I walk in to a customer and see 12 year old Winterm 1200 LE everywhere :)

Running Blazer 5.x ... :)


Shows how many stupid people there are. For a different use case "replace the desktop" making a 100% bet that they never a need to run anything locally. For a very controlled XenApp type environment makes sense.

Instead buy a cheap low form factor PC to do the same thing and you can manage it with existing tools and can choose from many models and maintain a very flexible strategy. Classic case of the blind following the hype.


It always irked me throwing away perfectly good hardware to replace it with slower, low power thin clients.

That's why I wrote ThinKIosk:

It allows you all the benefits of a windows device, offers power management options, can be centrally managed and also allows you to leverage your investment in your current hardware for as long as it lasts.


I agree... You can buy a small form factor PC (the same size of a thin client) with higher specs and low energy requirements for less than most thin clients.

Use existing deployments tools (e.g. MDT) and that cuts the costs down further by not having to have a new skill set and a new management platform.

If you have SA you can stick WinTPC on them, etc.

You wont feel so bad when you have to replace them after a few years because of a new protocol, etc.

Traditional Thin Clients SUCK!

To many retarded decision makers in the world...


Agree with appdetective - not sure why the hype continues... MS VDI costs alone should make $$ differential prohibitive - part of th reason Thin Desktop and Thin Browse have been succesfulfor us -    BTW, just completed a project where we we repurposed Thin Clients for a customer. A recently aquired thin client vendor told them they need to replace 3 year old devices to do "VDI Right" - wrong - repurposed the devices with Thin Desktop !


THX AppDetective and thinlaunch.... going to look into this - we have been told the same...makes our  investment in thin client seem worthless - what happened to 7 - 10 years????...Like the idea of using existing MS tools rather than Worst Device Manger


But why pay for a re-purposing solution when there are so many free ones out there?

Nice plug though...


Geez Dan - Give these guys a break... it looks like pretty cool stuff.Hhave you even looked at the products versus "free"? How about a positive comment for a change??


@basil, as Daniel said, why pay? Before you pay for thinlaunch, try ThinKiosk first, it's free and offers much more functionality. You wont regret it.

I'll be adding View support in the next few weeks.

if you have a further use case that ThinKiosk misses, I'd be happy to talk to you about it offline, my contact details are on my blog.


Ok... @AndrewMorgan 's ThinKiosk is ace! ;)

Quest do a good Windows shell replacement and a decent linux one. - Though Windows end-points will also be the best.

And yes, I've tried my fair share of paid for solutions.


two top reasons VDI and thinclients sales continue..

1. VDI is seen as cost savings measure, that will allow consoldation of support and control.

2. "everyone else is doing it" tied with its more secure


I can see repurposing existing hardware, but unless people are getting steep discounts off retail, I think there's no such thing as a $300 Dell that can run Windows and push out the video demands. Look at, and the cheapest PC they have is $300. It comes with a Celeron processor and 2GB of memory.

Perhaps that works, and I intend to get one and test it out.

There's one thing, though, that tips the scale towards thin clients for me - those $300 PCs come with Windows 7 Home. No SA, no VDA, nothing. So when talking about how MS VDI licensing requirements alone should guide you towards cheap PCs, remember that MS isn't giving this away for free.

Speaking retail pricing, $299 for the crappiest Dell, plus $99 for VDA means $398. If I can get a box that does the same thing from HP or Dell or nComputing that goes for $299 or less, I'm ahead of the game (since I have to buy VDA anyway).

AND, I don't have to worry about managing twice the number of Windows instances.

AND, I can use their PC to Thin Client conversion tools to leverage existing hardware in the meantime.

It's a tough call, because I like the Windows tools that Mike and Andrew have, but If I'm multiplying those numbers by thousands of endpoints, I think I'm coming out on the side of thin clients.

Maybe I'm missing something? Lock-in? I used to worry about that. HP solved it with their low end SoC's, but they are low-end. I just don't think people switch between XD and View that much to make it a huge deal.


@Gabe The idea behind ThinKiosk was really to allow you to recycle your current pc's as Thin Clients until the VDI technologies mature. So in that scenario, all you are really looking at is the VDA license.

Particularly if the customer is an SA customer with access to WIndows Thin PC.

I'm not favourable to either scenario any more than the other, ThinKiosk delivers benefits to both models ;)