Why hypervisor technology is foundational to critical automotive embedded systems

Elektrobit’s Kai Lampka and Adam Lackorzynski from Kernkonzept discuss how the development of multicore ECUs can be mastered inside future vehicles


View the webinar recording:

Download the slides:

Consumer electronics oriented multicore-based system on chip (SoC) designs have entered the automotive market to satisfy the higher computing demands foreseen with future in-car applications. Hypervisor technologies offer the opportunity to consolidate both existing and new software stacks in a single platform, thereby preserving investments already made in a car’s software base.

Whilst this trend has targeted mainly human-machine interface (HMI) related applications of the cockpit, the road to a completely centralised and non-closed computing infrastructure inside vehicles, possibly containing open source software, remains open. Unlike consumer-centric multicore SoCs as known from smart phones, industrial embedded SoCs need to adhere to strict timing, power, temperature and safety requirements on top of common security constraints, making the development of high-performance ECU such as in-car applications servers a challenging mission.

In this 60-minute webinar, Kai Lampka, Architect, EB corbos Hypervisor at Elektrobit, and Adam Lackorzynski, Chief Architect at Kernkonzept, investigated a concrete virtualisation layer and device sharing technologies, offering a prime example of how the aforementioned constraints can be addressed, and how massive multicore ECUs can be mastered inside future automobiles.