Is There a Market for Standalone Hypervisors?
The commercial market for mobile virtualization solutions was highly limited even before General Dynamics’ acquisition of Open Kernel Labs last month. Now, with several recent reports indicating that IBM is close to reaching an agreement to buy Red Bend Software, the commercial market may disappear entirely. But is this really that surprising?
For years, suppliers of mobile hypervisors have struggled to effectively communicate the value proposition of their solutions. Initially touted as a technology that can enable reduced bill-of-material costs, mobile virtualization has more recently been evangelized as the way to allow mobile devices to securely support multiple personas. In either case, device manufacturers failed to significantly support either premise – at least in terms of making investments in commercial solutions. As a result, revenues did not scale and leading vendors struggled to realize significant growth.
On the embedded side, the majority of the market is not based on standalone hypervisors. In this segment, virtualization has been primarily instituted as part of a broader runtime platform. Embedded virtualization solutions from vendors such as Green Hills Software, LynuxWorks, SYSGO, and Wind River heavily leverage – and in some cases are completely integrated with – each company’s flagship RTOS platform. Consequently, the success of these vendors has been driven in large part by their solutions’ ability to enable a guest operating system (Android/Linux or Windows, for example) to run alongside an RTOS in a multi-OS environment. Clearly, this value proposition has gained much more traction than the previously described mobile virtualization use cases.
So what do we see as the market moves forward?
According to GD’s press release, OK Labs will deploy its OKL4 Microvisor in secure mobile devices (for civilian, government, and military use) and automotive in-vehicle infotainment systems as part of the GD Broadband business unit – presumably within both internal and commercial opportunities. However, given the historical difficulties in monetizing mobile virtualization, we believe it may be only a matter of time before GD completely internalizes the use of OKL4 technology.
Of course, IBM’s intentions for Red Bend’s virtualization technology have not been publicly discussed, and the storyline is slightly different. Red Bend’s VLX product line (which the company acquired through its 2010 purchase of VirtualLogix) is a small fraction of Red Bend’s total business and is not likely to be the driving factor behind IBM’s interest. A more plausible scenario is that IBM is most focused on Red Bend’s mobile software/device management solutions, which are particularly attractive due to the rapid expansion of the BYOD paradigm. In fact, we suspect that it is possible – perhaps even likely – that IBM has no interest in VLX at all.
All this in mind, we expect that the commercial market for mobile and embedded virtualization solutions will begin to shift more heavily to traditional embedded markets. A large portion of this business will be driven by deployments within device classes utilizing a multi-OS environment in which Android or Windows powers the user interface with an RTOS controlling any critical/deterministic processes. Aside from smaller, specialized vendors such as Real-Time Systems and TenAsys, a majority of these solutions will be integrated runtime platforms as opposed to standalone hypervisor offerings.VDC will investigate these and other trends in our upcoming report, Virtualization for Mobile & Embedded Systems, from our research service Strategic Insights 2012: Embedded Software & Tools Market. Please contact us for more information.