The gating factor is acceptance by the OS vendors to include the technology as part of their core offering, which he thinks will take some time to occur. RIM's Goguen says RIM is not opposed to the idea of virtualization on the BlackBerry OS; Apple, Google, Hewlett-Packard's Palm division, and Microsoft declined to comment.
Although I like the concept, I'm not as optimistic that the major smartphone vendors will come around. Apple has made it clear it wants no alien apps on iOS, so far banning Flash, Java, and Adobe AIR; the only options are native and HTML5 apps. RIM's Giguen said he thought other technologies could fill the bill, so he didn't expect to see much enthusiasm for client virtualization, but RIM nonethless supports Flash Player, Adobe AIR, and Java on its BlackBerry OS (Adobe still hasn't delivered the first two, though).
Google has pooh-poohed Java, complaining about its weaknesses, and bought the Java-based Android instead. HP may have no issue with alien apps, given it is letting Adobe port Flash to WebOS. The two remaining major mobile platform vendors, Microsoft and Nokia, have supported Java already and thus may be open to the client virtualization concept as well, though both are fading away as relevant in the mobile space.
Under-the-hood virtualization. Another virtualization approach in the works is the use of a hypervisor between the device and the mobile OS(es), apps, and thin clients -- what's called Type 1 virtualization. OK Labs has deployed this approach on the Motorola Evoke cell phone, which can run both Linux and BREW applications but presents a unified interface to the user. The hypervisor approach is more secure than the Type 2 virtualization approach familiar to most users (where a virtual machine runs on top of an operating system, such as to run Windows on a Mac), says Rob McCammon, vice president of product management of OK Labs, because it has exclusive access to the device's processor, so can ensure separation between any operating systems and apps running on it.
McCammon says that device makers, cellular carriers, and even mobile OS providers are showing increased interest in the concept of such Type 1 mobile virtualiation. Enterprise IT is also expressing interest in the concept, he says, because it would allow secure, consistent management of devices across multiple platforms, which would support the strong "bring your own device" trend without compromising necessary IT control. Plus, it allows the deployment of a single app across multiple devices, assuming they share a primary or secondary operating system.
Multiple technologies combine. It's likely that IT, users, and vendors will combine multiple approaches. Certainly, sandboxing and cloudsourcing will be widely used, given their universality. If VMware succeeds in getting most or all major mobile OS vendors to adopt its embedded virtual client approach, client virtualization will get strong IT support for its enablement of cross-mobile applications, strong security, and cross-mobile management. Ditto if OK Labs' approach becomes common.
The thin client and VDI approaches probably will have the least uptake, given the complexity of setting up terminal services and the unsatisfying experience of running desktop and server apps on mobile device's constrained UIs. But for companies already deploying thin clients for desktop users (usually large enterprises and government agencies), implementing thin clients on mobile devices -- especially slates -- is a no-brainer.
Beyond unification: When the smartphone and PC merge
Both Citrix's Fleck and EMC VMware's Krishnamurti see the unification of personal and business on the mobile device as the first step to supplanting the PC. Given the processing capabilities and networking ability of mobile devices, why would users need a PC? If mobile devices can dock (likely wirelessly) to monitors, keyboards, mice, projectors, and the like, why have a separate PC?
Fleck says he already sees enterprises where mobile workers rely mainly on iPads running a mix of cloud apps, local apps, and a thin client to Windows. Today's iPad, iPod Touch, and iPhone support keyboards via Bluetooth, network access via Wi-Fi, and monitor/projector output via a VGA cable, he notes, so much of the needed connectivity to be dockable already is in place. The creation of Wi-Fi-enabled or UWB-enabled monitors and adding iOS 4 support for mice could take Apple's devices the rest of the way.
Slates and smartphones from other providers have similar capabilities -- most also support USB for physical connection to printers and other local devices. Local memory might be a gating factor today for iOS devices, but over time the built-in memory will only increase. And most competitors already have slots to add more storage.
Naturally, this is easier said than done. The truth is that there are very few Bluetooth keyboards in use at most companies, and there are no wireless monitors as yet (and imagine what happens when you unthinkingly pick up your iPhone to answer a call when it is tethered to your screen via a VGA-out cable). But the components necessary are not expensive and can be brought in through natural hardware refresh cycles once they become available. And we've seen how fast enterprises can adopt new technology when users, management, and IT are all excited.
"This will fundamentally change how enterprises think about computing," says Krishnamurti. "What does a PC client really mean -- is it tied to a physical desktop or not?" The answer is no. As mobile devices become the computer you always carry for personal and business use, and as virtualization, cloudsourcing, and thin clients push much of the computing to the network, you won't need a separate PC. It'll be a whole new world.
This article, "When the business and personal smartphone collide," was originally published at InfoWorld.com. Read more of Gruman et al.'s Mobile Edge blog and follow the latest developments in mobile technology at InfoWorld.com.