Because all major chip vendors push their built-in hardware virtualization extensions, virtualization in the embedded space is on the rise. However, virtualization itself has been a well known and well established technology for over two decades. IBM started to use virtualization for their mainframe computers in the early 80s, whereas the theory of paravirtualization is even older.

Virtualization can be used to solve a lot of the embedded industries´ problems. Hardware obsolescence, software migration, encapsulated software for safety and security reasons or utilizing today´s computing power more efficient are just some of them. But the built-in virtualization support of modern chips does not offer a complete virtualization solution. It is up to the software on top to utilize this virtualization support is an effective and transparent way. Paravirtualization is the most common and widely used approach in the embedded industry.

Paravirtualization requires a slight adaptation of the hosted environment to run on the virtualization layer, often refered to as hypervisor. A hypervisor can host a full featured operating system (OS) like Linux or Windows, a real-time operating system (RTOS) or a runtime environment (RTE) like Java or Ada. The small adaptation of the hosted environment, or guest environment, leads to a much better performance compared to a full hardware virtualization or emulation.

A paravirtualization real-time operating system (PVOS) is the next step for virtualization technology in the embedded space. A PVOS combines the virtualization capabilities of a hypervisor with the deterministic behaviour and possibilities of a modern real-time operating system. Therefore, real-time tasks can be directly created on the PVOS itself, so the PVOS does not require hosting another OS, RTOS or RTE. Applications can be directly developed on the PVOS.