Citrix close to acquiring RingCube for user personalization? (Our analysis of hypothetical deal.)

An interesting rumor popped up in the past few days that Citrix is close to acquiring RingCube, maker of the vDesk VDI user personalization software.

An interesting rumor popped up in the past few days that Citrix is close to acquiring RingCube, maker of the vDesk VDI user personalization software. Doing so would bolster Citrix's capabilities around user personalization for XenDesktop and XenClient, as well as giving them a decent play in the user installed applications space.

I've always loved RingCube's core technology. In the past I've said that they deserve an "attaboy" for putting together some cool stuff, but that I was less confident in their ability to survive in the market. They have great technology, but I'm nervous that the overall desktop virtualization market is developing more slowly then they'd need to survive. I wonder what their cash burn-to-actual sales ratio is now. (RingCube isn't unique in this predicement. I'm also nervous about Virtual Computer, MokaFive, and Wanova.)

This rumor came in via a few channels, and yesterday I called several folks who've been solid sources in the past for M&A activity in this market. Multiple people I spoke to produced the name "RingCube" without prompting, so I'm going to say the likelihood of this is fairly high. I also noticed that RingCube's CEO Pete Foley just added a new LinkedIn connection to Caroline Ghesquiere, a strategic development analyst at Citrix. Her LinkedIn job description reads "Support all strategic and corporate development activities including mergers and acquisitions, strategic investments and technology license activities." I recommend that you follow her new connections on LinkedIn if you want to keep tabs on what Citrix is up to. :)

If you're not familiar with RingCube, they have a product called vDesk that essentially "spawns" what appears to be a fully isolated Type 2-like virtual machine that runs on top of an existing copy of Windows. But unlike VMware Workstation or Microsoft Virtual PC, RingCube vDesk leverages the existing files (EXEs, DLLs, etc.) of the base copy of Windows to spawn the "VM." So instead of needing a multi-gigabyte VHD or VMDK to boot the VM, the entire vDesk package can be as small as 30mb. (VDesk leverages the same savings when it comes to memory, able to leverage the existing host OS files that are already loaded into memory instead of requiring a full GB or two to boot the VM.)

While the obvious advantage of vDesk is that you can boot an isolated VM based on just a few tens of MB of disk space and a few hundred MBs of memory, the disadvantage is that the version/patch/build level of the VM can only be the same as the host. Also, obivously, it only works with Windows hosts.

RingCube initially launched vDesk a few years ago as a solution for "local" or "offline" VDI, with the idea being that you could deliver a fully bootable VM in just 30MB. (Again, assuming that it was ok that the VM OS version and patch level would match whatever Windows desktop it was booting from.) But when they announced Version 4 a few months ago, RingCube changed their strategy, instead selling their product as the ultimate user personlization environment. (Imagine the isolation of side-by-side VMs without the overhead.)

RingCube's cofounder & CTO Kiran Kamity explained vDesk 4.0 at Citrix Synergy earlier this year in this 4-minute video interview.

What would Citrix do with RingCube?

Maybe the rumor of Citrix buying RingCube is just that: a rumor. But let's imagine that it's true. What would Citrix do with RingCube?


Obviously their first choice is to add vDesk to XenDesktop to enable a more full user personalization capability. This is something that Citrix is absolutely missing today. The closest thing they have is "Citrix Profile Management," (formerly called "User Profile Manager"), a capability Citrix got when they bought the Sepago Profiles product back in 2008. While it's a fine product for user profile management, it's *just* that--a fine product for user profile management. But in order to really enable users to customize and do whatever they want, you need a product that isolates and/or virtualizes the entire user environment, not just the Windows user profile. (After fifteen years, we all know that roaming profiles don't capture everything.) So buying RingCube could immediately give Citrix a "more than Profiles" capability. (Maybe RingCube + Shared Master Disk Image will avoid the TS versus VDI paradox?)

This is also great since a lot of Citrix's core XenDesktop capabilities are built around the concept of shared master disk images. (For example, XenServer's new IntelliCache functionality only works with shared desktops, as does XenClient's "Cached Image Mode.")

Second, we now know for sure that VMware View 5 will include a new profile management capability (which VMware got when they bought RTO Software last year). It looks like the capabilities of View 5's profile management will be very similar to Citrix's current Profile Management offering, so buying RingCube will enable Citrix to claim that they have better and more complete user personalization than VMware.

And of course, now that AppSense is in the spotlight after having received $70mm in funding from Goldman Sachs, Citrix buying RingCube can help them feel less bad about not buying AppSense when they had a chance a few years ago.

Potential Downsides?

I would imagine that Citrix can pick up RingCube fairly cheap. So as with previous lower-cost desktop acquisitions like Aurema and Kaviza, what has Citrix got to lose?

Longer term we have to keep in mind that RingCube can only help Citrix with Microsoft Windows ("Big D") Desktop-based user personalization. And while that's inline with everything that Citrix is doing now, it will have to change at some point. So can Citrix extract enough value out of RingCube in the meantime? I guess it depends on what they pay for them.

Join the conversation


Send me notifications when other members comment.

Please create a username to comment.

Not sure it will append except for a small amount of money... Reason why : how will they monetize it ? They will need to give it for free in one or any of their XenDesktop edition... Even if it makes sense technically, if they are not pushed (obliged) by the market because of market share threat... and there is no serious threat...


This would be WAY cool.  The only thing right now preventing me from using vDesk is that my company (large) is unwilling to base an architecture on a product offered by such a small company.  Though we use (and would likely CONTINUE to use AppSense for high- end Profile Management (it scales) and also App entitlement, the layering technology offered by vDesk makes non-persistent images viable  for a much larger number of use cases.  As is stands, there are far too many application "one-offs" to install them all on a master image, or use App-V or XenApp to deal with them.  vDesk would allow users to have their little specific apps run on their own .VHD with minimal effort.  Speaking of minimalism, the vDesk approach to deploying this technology is simplicity itself.  Is this were a standard feature of XenDesktop (and supported by the venerable Citrix) I would hate to be in a features war on the side of VMware View.  This would open up a whole range of new users to the VDI platform who were previously not considered "low hanging fruit."   Come on Citrix, and sign that check !!!


@jklincewicz : will you pay an extra to Citrix on your current XenApp/XenDesktop per user/device/ccu you have to get it ?


vDesk is now @$60/seat (about.)  If I could satisfy 4,000 users instead of 400, you BET I would pay extra for that.


And anyway, since Citrix paid like a half a BILLION for XenServer, which ended up being a "freebie" feature of XenDesktop, I'd say there is certainly a precedent, no ?


If it ever happens, it would be a great addition to the enterprise edition. I sure hope they don't consider having it exclusively for platinum, where it will give platinum more consideration but in reality it is an enterprise required feature.


This IS pretty interesting. This is consistent with my long standing prediction that hardware virtualization is the wrong solution. Virtualizing at the OS layer and App Layers is fundamentally more scalable and sustainable. The efficiency of OS virtualization is an order of magnitude higher.

Now that everyone is becoming aware of the complexity, cost and inefficiencies of VDI the first vendor to market with a workable solution that virtualizes at the OS layer will have a huge competitive advantage!!


Most Citrix customers are wishing for something like this from Citrix. The only apprehension is hoping they won't be forced to go from Advanced Edition to Enterprise Edition to get it. Or, imagine having to go from Enterprise Edition to Platinum Ediion to get it. All in all, It's easy see Citrix integrating with a product such as RingCube. It would take Citrix two or more years to develop it themselves (way too late and would they get it right?!) as opposed to 4 months to acquire/absorb it into their own Xen product line. If Citrix buys RingCube they should pass on their savings on development costs to thir customers instead of charging their customers more for it. I'm just sayin'...


I think as noted by jklincewicz  this shows that user personalization is not "solved" by a profile. And that VDI (despite its promise to be better than TS/RDS) is basically just a flevor of TS and its potential user base with out some type of decent personalization layer on top of it. Also I see he has noted that it would allow "users to have their little specific apps run on their own " This is key when we talk about centralizing desktops for users that could not use a shared RDS type environment.

As you can see he would pay for the cost to get satisfy 4000 users vs just 400. And this key. To me this is they types of changes needed to make VDI worth the effort.

Anyway, its good to see some of the "big boys" playing in this space. It starts to validate the user complaints about VDI and what they really need to move forward.


App compat is not going to get solved with OS layering until MS supports it in the core OS. Sure Citrix can buy this, but will take years to mature for app compat and many more problems will be uncovered. Once again a niche play that is very typical of Citrix.

I also maintain as I have always said UIA is a stupid use case. Nobody in IT is going to allow it and provide a service level. There have to be rules, or simply as Brian states, use a 2nd machines and do what you want. Wrong use case to solve for.

Also I don't see how this all gets lumped in with personalization. RingCube provides an app layer not granular personalization or user based managed, so a different problem. A major flaw in peoples understanding.

If Citrix buy RingCube it will be a feature of all XenDesktop editions because how else are they going to perform updates to the layered images.


Consolidation, consolidation, consolidation.

Since joining Cisco 8 months ago, VXI (Cisco's desktop virtualization solution) has gone from not being in the top 10 to the now the #1 most requested topic by customers globally in our customer and executive briefing centers (CBC & EBC).

One of the issues that CIOs and IT executives continue to raise in these CBC/EBC visits to Cisco is the importance of end-to-end leadership AND the fact there are TOO MANY POINT VENDORS to deal with causing risk and complication to their VDI projects.

All of these startups who are doing really cool innovation (like my former employer RingCube) need to be acquired to help the growth and maturity of the overall industry.

If Citrix buys RingCube and hopefully VMware buys Unidesk and AppSense & RES both get acquired by larger players, this will ultimately be the best thing for customers, partners, and the health of our industry to grow and mature into a mainstream offering that's common place in corporate environments.

Personally, I hope it happens and continues to happen. Kaviza, RingCube, AppSense, etc.  Acquire, acquire, acquire...



@dougdooley lets please pray that not all get bought. The legacy vendors like Citrix, Cisco, Microsoft represent old schools of thinking, slow to innovate, idiot leadership and have a view on the world that results in marketing hype, useless execution and forced marriages to old technologies. We need the smaller companies to get stronger and move beyond marriages to big brothers and push the envelope with new ways of thinking and delivering FASTER that the other idiots are not going to make things happen due to useless leadership.  



I understand the point about User Installed Apps and the policies that need to drive them. But in small shops (lets say 1000 seats or less) these are very very common. While not ideal, they are a reality, thus the drive you see from vendors to support them... the customers are asking for them.

Now this may just point to a different a problem. And in my view the core problem is that IT is unable to move at the speed that their user is. They need something on their desk now. IT can package, regresssion test, deploy etc in time. And of course this now has to go into full lifecycle mode for updates and what not...

Basically the user is moving too fast for the IT shop. One of my favorite stories about Dell was that I was an employee for less than 24 hours before I had admin rights on my laptop and was installing the apps I needed. Sure there were centralized apps I could "get", and I did. But mostly I got my personal apps that I used everyday. Everything from Tweetdeck to snagit, to gotomeeting clients, etc.

Anyway. Your point is well taken and the UIAs. But I think you would be surprised by the percent of organizations that allow it on a signifigant number of desktops.


@Ron if users want to do what ever they want then its not reasonable to expect IT to support all use cases. Perhaps its user supported apps- nothing to do with IT. If a service level is expected there must be some rules or the thing breaks and a cluster F is created. Most people allow it in small orgs because IT adds no value. Even if enabled I doubt the app compat issue will make it a main stream use case for these people. They are better off saying go virtual and do what you want locally. The real thing to manage is the data on virtual and local....

That said Unidesk is one of the vendors I would like to see keep pushing forward. I wonder why you don't do more with the user virt guys and create something really powerful together that would school Citrix, VMW and MS.


I'm really not as concerned with the user-installed apps as I am with the hundreds of custom apps (and their attendant libraries, .dlls, controls etc.)   Order of install can still be an issue (which we are looking at APP-DNA AppTitude to help mitigate.)  In a decent-sized environment (for a 25 year old company) there are lots and lots of legacy apps that just build up, and are still mission-critical to SOMEONE !!  I would personally have no issue with a small company (when I bought my first CBASIC compiler, Microsoft had three people including a secretary) but old-school IT folks are risk-averse.  I don't imagine Citrix can screw this up any more than they did Ardence (which is actually better in SOME ways) since they bought it.


After reading RingCube's site it is clear that have re-focused on the user workspace, including apps...

So what happens to RES, APPSENSE AND UNIDESK if Citrix pops on this?


@jklincewicz - I have to say, you write with the "voice of the customer"

In the last few months,  meeting dozens of CIOs of large Enterprise banks, healthcare, and government organizations - what you're saying is exactly what we keep hearing from them on a weekly basis.

On one hand those IT leaders request/demand, incredible amounts of innovation (e.g. they want all of Cisco's UC, voice, video and TelePresence software to be completely hardware/OS agnostic and run using HTML5 only - technically, that's incredibly difficult - doesn't matter if you're a startup or a big company), but when you ask them how many applications that are "business-critical" / "mission-critical" that only run on Windows (XP, Win7), the answer often comes back in the number between 2,000 - 12,000 apps of which a sizable portion are custom, home-grown apps.  Ask them about their plans to move those apps to the "cloud" / non-Windows platforms, the answer is either "no funding" or "no immediate plans" to do that in the next few years.

Compared to the consumer market, the rate of technology innovation consumed AND the appetite for risk is just not the same in the Enterprise market. The risk-aversion issue is real. And virtual desktop projects are directly run or funded by C-levels / VP-level IT execs in big companies.  And it's often their careers that are made or broken by these highly visible virtualization projects.

Thanks for sharing your comments. Spot on.



@dougdooley Custom apps many of them or old stuff that nobody has any idea about anymore. So you face a choice. Risk trying it in a layer (dept layer as per Brian's tweet earlier today). Would I really bother? Let me think about the choices.

1) Old or custom app requires effort testing in a new layer technology with no management, proven scale and app compat. Personally I'm likely to pass under the pressure of a migration. Maybe a futures technology. 2014 in time for Windows migration, seems like a huge career risk.

2) If one seems too risky, am I better of trying to figure out using application virtualization and pay MS MDOP fees? I'll probably pass on this also given MS is going nowhere in this space unless there is clear win and I can or have sucked up the MDOP cost. Less risky than option 1.

3) Say F it and simply stick to 1-1 like most people do or just go RDS, get stuff in the data center, solve real problems first and then start to get to better management. Along this simpler path I can probably get some app virt done, perhaps some user virt if I want to use citrix flexcast models vs. just VDI and build some momentum. Least risky path but also most costly due to storage unless I stick to local storage as well and maybe the promise of storage iO optimization, UCS etc become more real over time and deliver stable, high quality ROI.

All options to consider and no solver bullet. Layers is a technology that has a long time to mature IMO.


@Steve:  vDesk can do a modicum of Profile Management, and would certainly suffice for smaller installations.  For enterprise-class platforms, AppSense, RES, Liquidware ProfileUnity etc. are much more feature-rich.  

@dougdooley:  After a dozen or so year sin vendor /SI Presales, I am ducking back into being a customer .. where the end of a quarter is just another day (sigh..)  It does give you a different perspective.  When there has been no history of serious ITIL (Release Management, exhaustive testing etc.) apps are the biggest hurdle to VDI.  IMO, layering gives you the best of both worlds compared to sprawling 1-2 persistent images vs. one-size-fits all "master" shared non-persistent.  It is relatively easy to roll back a single VM if an install fails vs. trying it out on a shared image and affecting a large user base.


Oh, yeah ....Cirtix DID buy Sepago which competes in the profile space... just not very well. it makes the '"checklist" of features but you don't want to use it for an enterprise roll out.  


I think it's worth pointing out that vDesk VDI Edition is a totally different beast compared to the standard vDesk offering that most of this discussion has been based upon. The standard vDesk product allows the "spawning" of another instance of Windows like Brian mentioned in this article (and has been explained in detail by Brian in previous posts), however, the current release of vDesk 4 *VDI* Edition DOESN'T behave like its desktop cousin.

Instead, it uses what looks to be a file system filter driver to intercept all disk WRITES and redirect that to a user persistent disk (disk READS are a merge of the persistant disk + the master image). This user persistent disk is a mounted VHD, which is hidden and transparently overlayed with the system partition (kinda like the way file access with App virtualisation works).

This is pretty cool as it allows things such as AV updates, crash dumps, event logs etc to be retained between reboots, supports the SCCM agent in a pooled/master image environment (this is big) *and* also enables user installed apps. Of course, none of this would be useful if the user persistent disk was

invalidated once the master image was "recomposed" - which somehow RingCube seem to have under control with their agent.

So the result is it pretty much lets you treat pooled (single image) virtual machines as though they were dedicated (and duplicated) virtual machines – without loosing the storage benefits of a single image.

I've posted a few more thoughts and details of how it works, features and pitfalls here:



@Michael Gansel

"invalidated once the master image was "recomposed" - which somehow RingCube seem to have under control with their agent. "

This is exactly what layering via file system filters and the proper handling of the layers registry can allow. The trick here is dealing with conflicts. I am kinda worried about a layer like this on a block based image. Conflict resolution is hard enough without trying to do it between two different technologies hosting registry entries and files that may be in conflict. I think the BETTER path is that the base and applications are hosted in the same technology (a layer) and merged instead of simply layered... of course I may be partial since Unidesk does layering from the OS on up.

"So the result is it pretty much lets you treat pooled (single image) virtual machines as though they were dedicated (and duplicated) virtual machines – without loosing the storage benefits of a single image"

Thats the goal.


@Ron:    I don't see getting around using a tool like AppTitude to really characterize and remediate apps that need to co-exist on a single OS.  Even with SMS/SCCM there is always danger of breaking an image with out-of-order installs.  Of course, this is mitigated if you can do it to a single test image before incorporating it into a pooled "master."  Whether or not you are an ITIL fan, there is a reason the Release Management function exists in well-run shops.  I wish it ran in mine :(  <g>.


Brian predicted this one correctly:


Yes !  I just got confirmation from RingCube.  Good move, IMO..