Citrix Systems is moving aggressively to desktop virtualization with today's announcement of the new Citrix XenDesktop 2.0 products. Combined with other recent Citrix strategic moves, the world of PCs and applications delivered as services is soon to be flipped.
Heads or tails, both end users and those seeking to make a good living delivering business and consumer applications as services should win.
The slew of announcements come at Citrix's iForum user conference in Las Vegas, and quickly builds on the now-final acquisition of open source virtualization vendor XenSource, which Citrix picked up for $500 million in August.
Citrix XenDesktop combines Citrix Desktop Server, which uses the Citrix ICA (Independent Computing Architecture) protocol, with a virtual infrastructure for hosting virtual desktops in the data center based on Citrix XenServer.
The combo exploits dynamic provisioning to stream desktop images on demand from network storage based on the Citrix Provisioning Server (acquired with Ardence early in 2007). Citrix XenDesktop is due in the first half of 2008.
Using these technologies and approaches, entire PC desktops en masse can reside in efficient datacenters. And these are datacenters that can leverage and exploit: open source, virtualization instances of server runtimes and discretely supported applications, low-cost blades on standard hardware, automated provisioning and fail-over, and tightly managed and centralized operations. You'll get nice BI on how the apps and data are used, too.
In short, you get datacenters as the means to lifecycle delivery of apps, media, and web services dramatically lower TCO. It means a virtualized back-end utility-grid of delivery resources supports more of what has been a massive client-server money pit for going on 20 years. It means an applications delivery infrastructure that's actually under control, with declining total costs on energy and labor, that is flexibly able to deliver just about any environment, desktop, application, media and services.
When you combine virtualization benefits up and down the applications lifecycle -- with such functionality as back-end automated server instance provisioning -- you get excellent cost controls. You get excellent management, security and code controls. And you marry two of the hottest trends going -- powerfully low TCO for serving applications at scale with radically simpler and managed delivery via optimized WANs (NetScaler Web application accelerator) of those applications to the edge device.
A new type of ROI is now up for grabs, when you factor in datacenter consolidation, applications and middleware modernization, savings on labor, energy and real estate. And, golly, you'll be virtualizing Linux and Windows instances and serving up those platforms' applications as services right beside each other, running efficiently on the same highly utilized metal. See more on the cost and management benefits of virtualization runtime instances in a recent BriefingsDirect SOA Insights Edition podcast.
Incidentally, all of this augers really well for SOA -- discrete services can be supported and delivered this way too. And so they should be straight-forwardly composed and reused to build out flexible business processes. More on that another day.
For now, this end-to-end virtualization value that Citrix is quite close to assembling disrupts beyond the support cost-benefits analysis to include adoption and exploitation of new business models, such as subscription and targeted advertising for making such desktop and applications services very inexpensive or even free to those accessing them via a provider.
And the traditional channel is going to be shaken up, too. XenSource will reportedly soon announce an OEM deal with Dell and a resale support agreement with HP, says Internetnews.com.
Indeed, I expect that the Citrix solution set to begin to sell more among providers -- either outside of enterprises or internally with shared services and charge-back-based managed services bureaus -- than traditional IT departments. Citrix used to amount to a value of wrappering traditional apps with presentation services delivery to ease complexity and dealing with "problem" applications. Now, the value is about rethinking applications and their deployment lifecycles entirely -- and working toward the dual-necessity of improving the applications experience for users while dramatically cutting costs via simplified runtime environments and innovative economics.
Also on the disruption front, Citrix is now offering serious alternatives to virtualization market leader VMware, while also reming close (for the time being) to Microsoft and its forthcoming Viridian hypervisor. See more on the increasingly complex relationship in a recent BriefingsDirect SOA Insights Edition podcast.
So to take a step back and consider what Citrix is providing this week (and others will no doubt ned to step up to the plate with, too). We have the back-end and delivery benefits, catalyzed by virtualization. We have the managed delivery via Citrix's presentation, WAN optimization, and security services, etc. And the kingpin is the virtualized PC desktop as a service.
Yet because we're centralizing the delivery, we can also see how those services can be metered out on a per-user and per-service basis. So this enables the ecology of providers to offer comprehensive desktops and apps, as well as -- at the same time -- gives these service providers -- internal or external -- an economic means for charge-backs, managed services P&Ls, and subscriptions. You can make money. You can save money. You can do both.
Now, let's take it a step further. We can also inject and manage advertisements, training, knowledge-sharing, targeted links and content -- just about any relevant information in any media -- right into the actual presentation UI of the apps, media, services, content, etc. The PC desktop is fertile, untapped territory for these services. If I have to look at ads, I'd rather they have something to do with my work. Remember the BI benefits? By being centralized, the meta data on each user and apps use is there for the analysis and algorithmic associations. As users -- and their employers -- see the benefits of targeted content associated within applications and processes -- via display ads, links, RSS feeds, etc. -- they can empower the users, while also subsidizing the entire cost structure of providing the applications and services lifecycles.
This new monetization scheme would no doubt work differently for enterprises, small business, and consumers (and mobile users), but it could work very well through a variety of models. Again, you can make money. You can save money. You can do both. When you have centralized and managed serving of all the elements of work and play PC activities, the world is your oyster. You can innovate wildly.
All of this raises Citrix's profile dramatically, and makes for some interesting blue-sky "what ifs."
Think about "what if" Google had entree to this entire end-to-end apps delivery portfolio (and desktop virtualization jazz) and added it to its already heady SaaS offerings and massively effective targeted advertising arsenal. You could do Google web services, or Windows apps, or Linux apps (or green-screen mainframe apps!), all on a rich client or off the wire -- or whatever combo works best in the immediate circumstances. How about a full Google desktop. All of it is (or part of it) ad or subscription supported; all complementary with what's inside enterprises, and what's best acquired as services from outside.
So think about if Google were to partner closely (or even acquire) Citrix. Think about whether Microsoft could have the stomach for that. Imagine a bidding war between Google and Microsoft for Citrix solutions OEM deals (or the company itself). Or image Citrix remaining independent and playing to two nicely off of one another. Imagine that IBM might cotton to this as a way of getting in on the SaaS and ad-based models, while being applicable and amenable to the large enterprises. How about an IBM SOA desktop, a pallet for RIA mashups and Notes/Domino, too!
Sure, let Microsoft continue to dominate the applications development and deployment environment. And then use Citrix to provide for those applications and services simple, low-cost host and delivery alternatives (and multiple business models). All those VB developers are beavering away to create apps that you can better and more cheaply delivery via your virtualized and centrally provisioned environments. Microsoft subsidizes the applications creation, in effect, but only gets a portion of the pay-off on the deployments side in the form of Windows licenses for the virtualized server runtime instances. (But Microsoft begins to lose on PCs OS license, the Office license, and the cash cows go on a diet. Ouch!)
Meanwhile, let Google broker ads that can be injected (with permission) into the Citrix-powered services streams for all those applications. The cost savings for providing the apps goes closer to the degree of free for the business, subsidized by the cash from the targeted ads and hopefully useful content.
The myriad services providers and Internet providers adopt this all as the way to provide applications, desktops, content, media, and services off of the wire to small business, enterprise and home users at a compelling per month per user subscription rate. Those subscriptions can be baked into the triple or quad play of Internet, telephone, cable TV, mobile, and of all the needed or desired PC functions and applications. Talk about share of wallet!
I suppose I'd call that "Everyplay." Users can check off what they want, and its provisioned on the back end and readily delivered to the device (a low-cost converged device like the iPhone perhaps). And a home or small business could probably get all of what they need for less than $150 per month, with adds ons for beyond-basic services, (ringtones!), of course. Sounds like a business to me.
Yes, the Citrix strategy bears careful monitoring. The implications are really quite staggering. And this could happen sooner than you think.