Intel Study: Thin Client Competition Is Wide Open

Businesses are excited about client virtualization but no one model for getting there has a significant market lead over any other, according to a recent study conducted by Intel. That means there are plenty of opportunities for solution providers to advise companies exploring new compute models, says the manager of the chip giant's Emerging Model Program.

"The thin client story has been around for a long time, but there seems to be a resurgence of that story line in the industry. Yet when you look at the data, it's not playing out in terms of broad deployment," said Mike Ferron-Jones, discussing findings from Intel's "Emerging Compute Models and Their Status in the Market" study with ChannelWeb Thursday.

The study breaks out four "emerging compute models" and one older thick-client alternative, each of which has certain advantages and disadvantages, Ferron-Jones said. The "granddaddy" of thin-client (or more accurately, thin-terminal) technology is Terminal Services, represented by products like Citrix Presentation Server and Microsoft Terminal Server. The four emerging compute models studied by Intel are Virtual Hosted Desktop, Blade PC, OS + Application Streaming and Application Streaming.

While "virtualization" is often used in the popular press as a blanket term covering all thin-client models, the study separates the Virtual Hosted Desktop model from other, fundamentally different technologies. Intel is interested in those distinctions, Ferron-Jones said, "Because whatever model might be breaking away in the market will have profound impact on products we need to design for that market."

Among 705 "IT decision-makers" interviewed for the study, the most established alternative to thick-client systems, Terminal Services, scored highest in awareness of the compute model (96 percent); familiarity with the technology (84 percent); deployment in any capacity, including test installations (64 percent); and high-volume production installations (31 percent).

Interestingly, though, the study found that the Terminal Services model isn't being newly adopted by companies at nearly the clip that other, newer models are. While 64 percent of respondents said their companies have deployed Terminal Services in some capacity, just an additional 2 percent reported plans to deploy within the next two years. That 3 percent growth rate is dwarfed by a 27 percent anticipated growth rate for the OS Streaming (15 percent current deployment to 19 percent within two years) and Blade PC (26 percent to 33 percent) models; 21 percent growth for Desktop Virtualization (39 percent to 47 percent); and 20 percent for Application Streaming (30 percent to 36 percent).

"Everyone who wants it, already has it," said Ferron-Jones of Terminal Services, adding that while the data suggests that the newer models offer far more growth potential, "there's no breakaway winner in the market yet."

"If I had to pick a leader out of all of them, I'd say streaming as a whole is it, but that's a very tentative suggestion," he said, referring to the OS Streaming and Application Streaming models as technological "cousins" that could arguably be grouped in one bucket. Streaming solutions available now include Citrix Provisioning Server and Dell On-Demand Streaming Solution (OS Streaming) and Microsoft SoftGrid, Citrix Presentation Server, Altiris SVS, AppStream, LANDesk and Thinstall (Application Streaming).

Next: Citrix Spreading Its Bets Wisely

While no single alternative compute technology is running away from the others, one company has far and away the most diverse presence in each of the emerging models. Fort Lauderdale, Fla.-based Citrix has leading products in all the models examined by Intel, including Terminal Services where it competes with Microsoft for leadership in the space.

Citrix also challenges VMware's Virtual Desktop Infrastructure with its own XenDesktop Server in the Virtual Hosted Desktop model, where Microsoft is becoming a player as well. That's important, because according to Ferron-Jones the different emerging compute models each have their pluses and minuses, and are tailored for different IT environments. Citrix is a leading vendor in four out of the five categories studied by Intel, the exception being Blade PCs, where HP and ClearCube are named as the top OEMs.

All five of the models offer the advantage of centralized data security and management of applications, or entire images in the case of OS Streaming and Virtual Hosted Desktop, Ferron-Jones said, a top reason cited by respondents for their interest in alternative compute models. Total cost of ownership (TCO) relative to "typically managed rich desktops" is another plus, with significant, roughly equal annual TCO savings for Terminal Services, Virtual Hosted Desktop and the streaming models. Blade PCs, with their high acquisition and conversion costs, don't fare so well on TCO.

Going the Blade PC route also requires vendor lock-in for hardware and tools that might diminish its appeal for some, said Ferron-Jones. On the other hand, the model provides the benefit of a single hardware stack for validation.

Other areas of concern for adopters include user customization capabilities, where Terminal Services breaks down in comparison to the newer models, the compute-intensive nature of Virtual Hosted Desktop, and some reported inefficiencies with sequencing OS and application streams to clients in the streaming models.

But the biggest downside to all of the emerging compute models is mobility, or a lack thereof, said Ferron-Jones. Application Streaming is the only alternative that offers anything approaching the mobility many end-users require, and that's limited at best, say some critics.

"The downside is that you're tethered to the network. Sure, you can be wired or wireless, so there's some level of portability, but as soon as the terminal goes off the network, you can't work with the applications," said Ferron-Jones.

Because mobility is "the the other big trend that's rocking the world," he believes rumors of the thick client's death have been greatly exaggerated, at least for the foreseeable future.

"Sure, I think that wireless broadband connectivity is going to grow. And perhaps someday in the future, we'll all be walking around in a universal cloud of broadband coverage. But we're not there today. And we're a long way from there. In the immediate future, there's still a lot of opportunity to build up people's ability to be productive in low bandwidth or out of band," Ferron-Jones said.