Home Applications A First Look at Microsoft’s Client-Side Hypervisor and a Critical Look at their Desktop Virtualization Strategy and Desktop Virtualization Licensing Models

A First Look at Microsoft’s Client-Side Hypervisor and a Critical Look at their Desktop Virtualization Strategy and Desktop Virtualization Licensing Models


Written by: Guise Bule, CEO tuCloud Global

For the last few years, it was obvious that Microsoft did not really have a desktop virtualization strategy and they lumbered around for a while, trying to figure out which way the wind was going to blow, lacking the confidence to realize that if they themselves blew hard enough, they just might be able to make the wind blow their way.
We saw them take stabs at VDI, Terminal Services became Remote Desktop Services which of course fooled no-one, even Microsoft cannot convince the world that a server OS is a suitable replacement for a Desktop OS.
They realized early on that RDP just was not up to the task of serving server hosted virtual desktops over the WAN , a quick purchase of Calista and in a little bit of time, they managed to make something semi-respectable out of RDP 7, which is nowhere near as good as Citrix ICA/HDX.

United States Patent Application 20110010714

Inventors: Powell; Therron L.; (Redmond, WA) ; Anderson; Jason Michael; (Snoqualmie, WA)


The culmination of Microsoft’s Calista efforts will be the much-lauded Remote FX, but you can tell their heart isn’t really in it. They are giving the tech to Citrix to integrate into HDX.

I think this act signifies the end of Microsoft’s efforts to have anything to do with desktops that are hosted centrally on servers and delivered over the WAN. They are perfectly happy for Citrix to take the lead in these matters, to spend all that money on a tech they think is going nowhere.
We could already discern this from their licensing strategy, not only did they have a really poor VDI offering that was going nowhere, they also systematically sabotaged the efforts of the emerging desktop virtualization space to properly deliver Windows Desktop OS’s over the WAN.
In short, Microsoft hate the idea of cloud hosted desktops, this much we know, but how much do we know about where they are really going with the ‘desktop’? Not much at all really, mainly because they haven’t known themselves up until very recently.
You can imagine Balmer sitting there whilst his bright young things explain the concept of cloud-hosted desktops to him, whilst he holds a brand new HP Zero client in has hand that doesn’t run Microsoft Windows. I betcha anything that they didn’t like the sound of low cost zero-clients on users desks in offices all around the world, instead of high-cost fat personal computers running full Windows desktop OS’s.
It just wouldn’t sound right to them and I can imagine the idea scaring them more than a little, especially when they worked out application virtualization and the idea that one day not too far from now, the application layer will have absolutely nothing to do with the underlying desktop OS. In short, applications of the future will not need Windows to run.
So the question is, when the desktop becomes nothing more than a pretty thing for cloud hosted virtualized applications and cloud hosted data to sit on top of, what do we need to pay the Microsoft tax for? Well we won’t, because we will not need the traditional desktop OS anymore.
Scary stuff for Microsoft to have to deal with, you can understand why they just don’t want to deal with it and their knee-jerk reaction which resulted in the licensing fiasco we now find ourselves in.
Back in 2009, Gartner predicted that the hosted virtual desktop space would surpass $65 Billon dollars by 2013, but I can tell you that from where I am standing in 2012, it doesn’t look like we are on target to hit that figure, primarily because Microsoft is strangling the hosted virtual desktop market by (once again) using their weight to dominate proceedings.
I am going to ignore the story and explanation of VDA’s, VECD and other crazy Microsoft licensing taxes on hosted virtual desktops and come straight to the point in saying that in 2012, hosted virtual desktop providers are still not allowed to properly license a windows 7 desktop and rent say ten of them to a small business customer on a monthly basis, its nuts.
They do not allow us to sell hosted virtual desktops to the small businesses out there who would really benefit from being able to get rid of 15 of their fat Windows XP or even Vista PC’s and replace them with nice HP zero clients that let our customer access their Windows 7 desktops in the cloud. In 2012 there still does not exist a mechanism, whereby managed service providers can legally provide desktop OS’s in virtualized form which are deployed over the web.
This completely sucks, because not only as a business owner I am forced to turn away prospective SME clients all day, even worse, a lot of the time I also have to explain to them (time and time again) why I cannot rent them five Windows 7 desktops for a month or two, even though I have a beautiful award-winning, best-of-breed, large scale Kaviza/HDX platform that is perfectly designed to do just that. Sometimes I don’t think they believe me it’s so nutty.
I look at the hosted virtual desktop space around me and I see most of my competitors, struggling to deal with Microsoft licensing and all deploying Terminal Services/server desktops to their customers whilst pretending that they are real cloud hosted, Windows 7, virtual desktops. 
All we want to do is rent out Windows 7 desktops to SME customers who really want to rent them and Microsoft will just not let us. Instead Microsoft is smothering the space with its licensing regime and simultaneously forcing its inhabitants to use hosted virtual desktops that are basically just terminal services desktops with a lick of fresh paint hastily daubed on them.
80% of the HVD space is still deploying terminal services and slices of server OS, masked to look like a real Windows 7 desktop OS and all to get around Microsoft’s ban on the delivery of desktop OS’s in HVD form. 
I decided last year that I was going to have nothing more to do with terminal services and my company now focuses on deploying real virtualized, non-persistent, desktop OS’d such as Windows 7 to our clients with our award winning hosted virtual desktop platform.
Well hold on a minute you must be thinking, we just spent two pages learning about Microsoft’s crazy licensing and what you are not allowed to do, to be now told you are doing hosted virtual desktops anyway, even though it’s illegal? Herein lays the great hypocrisy.
Small companies who really need hosted virtual desktops are not allowed to use them but large enterprise sized organizations who give Microsoft millions dollars a year to purchase Volume Licensing and Software Assurance are allowed to use hosted virtual desktops.
In fact Microsoft make it easy for them by letting enterprise sized organizations leverage their existing volume licensing and software assurance onto out DaaS hosted virtual desktop platform for free. FOR FREE!   Consider that the main reason VDI isn’t getting traction in the space, is primarily because Microsoft licensing costs add too much money to the equation.
I just deployed thousands of hosted virtual desktops to a federal institution in a completely legit manner as far as Microsoft are concerned and the only ‘restriction’ on this, is that the organization has to report the number of additional virtual desktops they are using to Microsoft and they have to give me the license keys to activate the desktops.
But I cannot rent out five Windows 7 desktops for three months to a small company who could benefit hugely from using them.   The difference being of course that for the SME, there is just no Microsoft licensing model that works for them and typically they do not use volume licensing or software assurance and as such are not in a position to properly leverage desktop virtualization technologies, in order to add value to their businesses.
In other words, it’s fine for the big companies, but screw the SME market. 
The double-standards here are breathtaking but their thinking is clear. Microsoft haven’t got the balls to tell their enterprise clients that they cannot do whatever the hell they want with the millions of desktop OS licenses they are buying.
But with the SME market and the ecosystem of hosted virtual desktop providers that barely survive in that market, Microsoft has enough weight to distort the market in a major way by severely restricting the markets ability to prosper. We have heard this story before no?
Where has Gartner’s $65 Billion HVD dollars gone? Ask Microsoft. As far as I am concerned, they stole my slice of those billions right out of me and my competitors pockets.
What has all this got to do with client-side hypervisors you cry? Hold your horses, we are getting there and with this in mind I would like to introduce you to Microsoft’s desktop strategy moving forward, but before I could rattle on about a different kind of virtual desktop entirely, I first needed to fill you in on their schizophrenic approach to cloud based hosted virtual desktops. Now that we understand one side of their approach, we can use this knowledge to provide us with more clarity on the other side of their approach to desktop virtualization.
Yes folks, Microsoft finally decided that client-side hypervisors are the way forward and I must say that I do think that this is something that they are getting right.
Of course Microsoft being Microsoft, they don’t call a client-side hypervisor a client-side hypervisor and instead begin to talk nonsense about FAST BOOTING A COMPUTING DEVICE TO A SPECIALIZED EXPERIENCE.
If you do not know what a client-side, type-1 bare metal hypervisor is, go and read my explanation of the tech here and then come back when you think you do, because it all gets a little tricky right about now, mainly because I think that Microsoft may be trying to operate in some kind of stealth mode with this whole thing. I simply cannot fathom any other reason why they would not announce that they saw the type-1 as the way forward, unless of course they are trying not to burn their partner’s significant investments in VDI technology into waste paper. Yet!
If you talk to puritan client-hypervisor guys, they will tell you that VDI is dead and whilst I do not agree with this point of view completely, I do think that their arguments have merit.
Looking at my space from a global perspective, I can see immediately that large chunks of the world do not have access to high bandwidth internet connectivity, the kind we usually prefer our customer have when we migrate them across to a hosted VDI solution.
In fact, MOST of the World does not enjoy this and if we are honest, large chunks of our own Western countries still do not yet have great connectivity and this is an issue if you plan to migrate the world across to cloud hosted virtual desktops. 
When the concept of VDI and thin clients was invented at Sun Microsystems back in the day, they envisioned an Internet Age, a time when we would all have access to boundless amounts of network connectivity. We still are not there yet, in fact we are nowhere close and even worse I see the large telcos’s, the backward bastards, now talking about throttling out internet and charging us the consumers higher rates in order to access better connectivity and also going after the big content providers for money on the basis that their content is costing them money to deliver.
Did I say backwards bastards? Yes I did, because that is exactly what they are. I really wish somebody would slap these people around the face and explain to them that internet connectivity needs to be enhanced, upgraded and that money needs to be spent in order to do this. I wish somebody would force them to do this, because they clearly are not going to do this on their own.   Internet connectivity is like water, we NEED it and we need it flowing through our pipes as fast as it will come (especially in VDI), this of course is wasted on the telcos.
No incumbent entrenched and money motivated corrupt bureaucracy does anything that is not in its interests, not even when it’s in the interests of their own customers. When we see this type of behavior, we need to stamp down on its throat hard or we as a society will lose out to the self-perpetuating bureaucracies that dominate our society. But anyway, I digress.
So are we all in general agreement that the world is not quite ready for VDI in terms of the amount of network connectivity available to us? Yes?  
The type-1, bare-metal, client-side hypervisor. What a mouthful that is! From now on let’s agree to just call them CS hypervisors on the basis that a type-2 is a toy in comparison to the real thing.
So yeah, Microsoft is really starting to dig the CS hypervisor and this much is obvious from their patent application, some of which is plastered in the front of this paper.
Except, like I said earlier, they are not calling a kettle a kettle and instead they have dreamed up the amazingly original concept of a dual-boot system, which may be familiar to some of you because it is nothing new. We have all seen computers that let you boot into an alternate environment to do something like browse the web or play music, rather than fire up the OS to get access to a browser to do this.
Apple Mac users will be scratching their heads right now as we are not used to the idea of an OS taking forever to boot up like Windows users are and therefore the idea of a second environment which allows you to do simple tasks that your OS should be letting you do seems downright silly. 
No matter how much of a Windows fan you are, you have to admit that Windows is a dog when it comes to boot up times and that this doesn’t appear to have gotten much better as time has gone by.
It’s still a dog even when you have a blazingly fast machine running the OS as a guest under the latest hypervisor and their approach to deal with this is “Fast Booting a computing device to a specialized experience” which is code for, our OS is a dog at start-up, lets adopt a really old idea to fix this rather than be innovative or (horror of horrors) redesign our OS from the ground up so it works better. What happened to Longhorn? Where did you lose your way?
Take a look at the first illustration of their patent, the one that clearly shows a hypervisor virtual machine manager, sitting on top of PC hardware with a Guest OS on top of it. If that isn’t a CS hypervisor, then I don’t know what is.
That one picture tells us all we need to know about Microsoft’s desktop strategy moving forward.
The second image tells us that the system is designed to accommodate multiple guest OS’s on top of that hypervisor and also that the guest OS’s will be interchangeable and also perhaps they are thinking of service providers adding value to this mix, although I may be confusing the meaning of that second images when i see the words “Virtualized Service Provider”.
Thank you Microsoft for giving me a glimpse of your strategy and helping me see where you are heading so that my company can plan ahead.
I have already seen this tech before, tuCloud is a longtime partner of Virtual Computer, the splendid company that produces the wonderful NXtop, the only CS hypervisor that actually works on the market and has been used out there in production in the real world.
Anyway, when I saw that image, everything became clear to me. After a few conversations with some analysts, it’s obvious that these guys have been telling Microsoft for ages that the CS hypervisor is the way forward and in fact Microsoft have already acknowledged that Windows 8 will contain a CS hypervisor that isn’t Windows XP mode.
My advice to all of you organizations out there who are currently considering migrating across to Windows 7 is DO NOT DO IT. They are about to re-invent Virtual Computers work and call it Windows 8 and if you think migrating from XP or Vista to 7 is tricky, you aint seen nothing yet.
From this perspective, Windows 7 looks like a desperate scramble away from the Vista debacle towards something better (ANYTHING BETTER), but we can clearly see from their patent images and what we do know about Windows 8 that this scramble isn’t over yet.
I can kind of imagine a whole bunch of Windows 7 engineers sitting around after working for years on their baby, when the realization that desktop virtualization had just changed the game sunk in and them saying “Aiiiiiiiiiiiii” whilst they gnash their teeth. I would. 
Trust me, upgrading from 7 to 8 is going to take more than an upgrade disk. That’s a serious migration right there, but a very necessary one nonetheless.
In a way, the CS hypervisor is the ONLY way that Microsoft can go if they want to perpetuate their ‘dominance’ over the desktop space. I would argue that this dominance we all talk about is nothing more than a perception that has been enabled by our own thinking and perpetuated by some of the usual tricks on Microsoft’s part to achieve vendor lock in, but please don’t get me started as I have to finish this article one day.
If they want to keep the PC ‘fat’ and pretty much the way it is now in terms of how its licensed and how its managed then they must go down the CS hypervisor route, there is quite literally no other way for them and they are trapped by technologies which have emerged around them, taking them by surprise. Their engineers did not see this coming a few years ago and have been blindsided by it.
But at least they are adapting to this emerging technology and at least they are doing it the right way, even if it did take them forever to realize which way the right way was.
Microsoft want to keep PC’s on the tables in your office, but they also need to acknowledge that desktop virtualization means that we expect to be able to manage our OS’s in a much more effective and sensible way, the CS hypervisor is the best way for them to achieve this.
And with this in mind, let em tell you what I would do if I were Microsoft, as a way of explaining where I think they are going with this.
First of all, do not be fooled by this dual-boot nonsense. 
This is a Microsoft CS hypervisor in the process of being built and I for one have always thought that if Microsoft went this way, they would be dangerous. Dangerous in the sense that we may find ourselves stuck in yet another few decades of Windows, with little innovation and diversity in the desktop space. I believe in diversity and think that Windows and Mac OS X is not enough.
Their version of the CS hypervisor is going to be THAT good and here is what they (Or I) would do with it. Switching to MARKETING MODE.
Imagine a family computer in the den at your home, where each member of your family has their own virtual desktop on that machine. When you sit down to use it, you don’t have to deal with anything caused by the other users. Your son can download porn and cracked software as much as he wants, because no matter what he does to your desktop his desktop, no matter how many viruses or Trojans he gets, it’s not going to screw up your desktop.
You can have all kinds of virtual desktops, how about one specifically for online banking and used for nothing but? If that’s all you use it for, you know it’s not going to have any viruses or key loggers giving away your password details, because all you do is visit secure and trusted sites.
What about those viruses? Well we all know Windows is a virus magnet, so much so that sometimes I think that’s what it was invented for. We also know (the more tech savvy amongst us) that we lost the war a while back and all of our expensive AV does not protect us from these threats very effectively. AV is reactive rather than proactive and useless against zero day threats and botnet herds are getting bigger and stronger.
The solution to all of that is non-persistency on the desktop, when you log out your desktop reverts back to its natural state, wiping away any nastiness you may have picked up along the way and presents you with a fresh image upon login. Actually, that’s a little too advanced for Microsoft yet, so let’s just say for now, that if you get a virus, you can revert your desktop back to an earlier state when you didn’t (if your AV detects it) which is still pretty cool.
Need to fix your sons desktop because he broke it? Just wind it back.
The use case I describe above doesn’t even begin to talk about the awesome power that CS hypervisors give you from a management perspective. If you have to manage thousands of Windows desktops right now AND their applications AND support their users, I pity you, I really do. In fact you probably pity yourself; it’s a thankless job that can drive a sane man crazy.
But imagine if you can manage these desktops remotely, update just one of them rather than thousands of them and install apps easily to a golden image and have thousands of desktops instantly have the app installed on them. That’s powerful.
Users never back up, but why do they need to if whenever they make a change, that change is automatically backed up to a snapshot of the image held on a server somewhere (Azure??), without the user actually doing anything to make this happen.
Imagine if one of your field sales guys gets his laptop stolen, all you need to do is whip out a fresh laptop and dump his up-to0date desktop image onto the machine, then FedEx it out to him.
Starting to sound awesome yet? Because IT IS. Did I say they would be dangerous?
Here is something that’s awesome, the “No-Compromise Solution” which sounds grand and a little hard to believe, but is absolutely true. I first bumped into this one when I was deploying Virtual Computers NXtop into a very large financial advisory firm for their mobile financial advisors laptops. It goes a little something like this:
Imagine IT central could inflict upon you their dream desktop and how horrible that would be. Those guys don’t want you visiting Hotmail, or downloading things, they don’t want you having admin rights and believe me they think you are too stupid to handle admin rights.   In most cases, they don’t want you to be able to connect to anything but the corporate VPN or using the machine for personal use at all. It’s not a PS in the sense a Personal Computer should be.
Well now, IT central can inflict that desktop onto you easily and you will accept it, because as they give you that desktop, they will also give you a second Windows 7 desktop which we will call your ‘personal’ desktop. Even better, you can hotkey between the two in real time which is awesome.
All of a sudden we have a no-compromise solution, because the user gets what they want and so do IT central at the same time. When do you ever see such a no-compromise solution in real life?
Wherever we deploy this tech, the users and also IT central LOVE it, because everybody gets what they want. Happy faces on desktop users and IT central at the same time? Lovely.
Even better because it’s not a server intensive as VDI, the back end infrastructure is a fraction of the cost, reducing the TCO of the solution compared to VDI dramatically.
Now that is what you call a dynamite solution and this is exactly what Microsoft have in mind for all of their talk about fast booting computing devices into specialized experiences. It actually looks like they are trying to patent something Virtual Computer is already doing in that Virtual Computers NXtop are already providing fast booting into a specialized environment in that they allow you to boot into a remotely hosted virtual desktop and/or browser rather than boot up the guest OS’s residing on the CS hypervisor. 
Watch out Virtual Computer! I would contest this patent for sure if I were them. Microsoft have some cheek even trying to patent this, it’s somebody else’s idea and I think that’s partly why they are talking about all this specialized experience crap, it could explain their ‘steal’ tactic here.
Imagine doing something for years and then having Microsoft come after you because they patented your idea before you and that it hadn’t occurred to you to patent it, because you thought that’s where all desktops were going and so didn’t expect it.
I would be pissed.
Now that I have all my ranting out of the way and I have described where I think Microsoft is going and why, let’s look at some of the challenges Microsoft have ahead of them if they take this path, which they clearly are.
Challenge Number 1: Licensing!!! : My god how are they going to license multiple guest OS’s sitting on the CS hypervisor? Judging by their previous attempts at virtual desktop licensing, I would guess badly. But then again I am probably underestimating them here and using their strangulation of the HVD space to discern the way they will license this. I bet that however they choose to license this, it will go along the lines of “charge them as much as we can for as long as we can” and have some bullshit limits of the number of VM’s you can run under the Home Premium license or whatever.
They really need to learn to lighten up on their licensing, or at the very least stop their hypocrisy and double standards with VDI licensing. My advice to Microsoft is to “WAKE UP, STOP BEING SCARED” and let your users license your desktops properly in the way that will most add value to their own organizations, whilst they still want to use Windows. Their licensing strategy is causing much more anti-Microsoft sentiment than anything else and this sentiment is growing stronger daily.
Challenge Number 2: Hardware Compatibility!! : It has been mentioned to me that because Microsoft are Microsoft, they have access to all of the drivers and whatnot, which will allow them to easily overcome the hardware compatibility issues that they will face, but I don’t think this is true.  
Look at the Citrix XenClient hardware compatibility list (HCL) and you will see it’s not worth looking at, especially when compared to Virtual Computers HCL, in theory as Microsoft’s partner, Citrix should have an easier job at this, but they don’t.
If your CS hypervisor only works on a few bit os hardware out there, it will not be embraced as Citrix have discovered. VMware don’t even want to go there, hence their messing with type-2’s.
Also remember the last time you asked Windows to go find your driver when you plug something new in and it does? NEVER HAPPENED. Their driver mechanism is an embarrassing joke compared to Apples. 
I shit you not, I have never had an issue plugging anything into my Mac, and it goes and gets the driver every time if it doesn’t have it already. How can Microsoft screw this up so badly?
Challenge Number 3: Competition: By the time they get around to releasing it, others would have already blazed the way and the world may have moved on from the point where they want Microsoft to be their CS hypervisor and nothing but just another guest OS. Will Microsoft’s CS support Linux or Mac desktops? 
Will it let you migrate your VM’s over to XenClient or NXtop? I would put money on the fact it would not, because it doesn’t fit in with their agenda of perpetuating themselves as the dominators of the desktop space.
In fact I think they lost that title already and that the world and they have just not woken up to that fact yet. Give it a little time, our space is still a young one, feeling its way.
And to my competitors, to my customers and to all of you who really want to be able to use Windows desktops in a way that adds value to your business, I leave you this message.
Microsoft only gets away with their behavior because we let them. They are not strong, in fact they have never been weaker and this is especially true in the virtual desktop space, right now and in this moment.
I have lots more to write around this subject, but I need to stop this article now before it gets too long. I am sure to have made mistakes, if so I will eat my words.
So tell me if I am wrong, tell me if I made mistakes and why! Bring it on!
I love a good argument; it makes me feel alive 🙂
Guise Bule is the Fist of The North Star for the Desktop Superhero Alliance, the founder of the Virtual Desktop Forum and CEO of tuCloud.  Guise is a smart-ass blogger who likes to touch things that he shouldn’t and pick fights if he can, he is also an expert on desktop virtualization, hosted virtual desktops and cloud computing, even if he does say so himself.


DABCC DABCC.com, the world leader in sharing the finest Virtualization & Cloud news and support resources. #Citrix, #VMware, #Microsoft, #Mobility and much more! Brought to you by @douglasabrown & team!

Featured Resources:

Related Articles:


White Papers

    Application Lifecycle Management with Stratusphere UX – White Paper

    Enterprises today are faced with many challenges, and among those at the top of the list is the struggle surrounding the design, deployment, management and operations that support desktop applications. The demand for applications is increasing at an exponential rate, and organizations are being forced to consider platforms beyond physical, virtual and cloud-based environments. Users […]


      Download Commvault VM Backup and Recovery: end-to-end VM backup, recovery and cloud management

      Commvault’s ability to provide end-to-end VM backup, recovery and cloud management creates a significantly better way to build, protect and optimize VMs throughout their lifecycle. Our best-in-class software for VM backup, recovery and cloud management delivers a number of significant benefits, including: VM recovery with live recovery options; backup to and in the cloud; custom-fit […]

      On-Demand Webinars

        What’s Going on in EUC Printing – A Technical Deep Dive!

        The IGEL Community and ThinPrint invite you to watch the following technical deep dive webinar. The agenda is to technically bring you up to speed on what’s going on in the EUC Printing space today along with a deep dive into new methods, technologies, printing scenarios and a discussion on why printing still matters. You […]

        Latest Videos

          Views All IT News on DABCC.com
          Views All IT Videos on DABCC.com
          Win big $$, visit ITBaller.com for more info!

          Visit Our Sponsors