Contents

Hypervisors and virtualisation

How one SoC runs several isolated operating systems of different criticality — and what a hypervisor must guarantee for safety and security.

Why it exists

Consolidation puts infotainment, vehicle functions and safety supervision on one chip. They have very different criticality, update rates and suppliers — and must not disturb each other.

What it is

A type-1 hypervisor runs directly on the hardware and partitions CPU cores, memory and devices into virtual machines, each with its own OS (e.g. Android, Linux, QNX).

Type-1 hypervisor stack
Virtual machinesEach runs its own guest OS
HypervisorPartitions resources and enforces isolation
SoC hardwareCPU cores, memory and devices

Guest operating systems run in separate VMs above a type-1 hypervisor that runs directly on the hardware.

How it ensures isolation

  • Spatial: each VM sees only its memory; the SMMU restricts device DMA.
  • Temporal: CPU time and sometimes cache and memory bandwidth are reserved per VM.
  • Communication: VMs talk only through defined channels (shared memory, VirtIO).

Best practices

  • Keep the hypervisor small; use a certified one for safety-relevant partitions.
  • Run the highest-ASIL functions on the safety island, outside the hypervisor.

Common pitfalls

  • Shared caches and memory bandwidth breaking timing isolation despite separate cores.

What it means for your discipline

  • HPC & application software Core

    Designs VMs, resource allocation, inter-VM communication and device sharing.

  • Functional safety Core

    The hypervisor must provide spatial and temporal isolation (freedom from interference); certified hypervisors exist.

  • Cybersecurity Core

    Isolation keeps a compromised infotainment VM away from vehicle functions; the hypervisor is part of the trusted computing base.

  • Hardware & silicon Partial

    Relies on virtualisation extensions, SMMU and cache/bandwidth partitioning in the SoC.

  • Software update & OTA Partial

    Plans update and restart boundaries for guest images and the hypervisor, including version dependencies and effects on other VMs.

No direct stake: E/E architecture · Base software · Networks & communication · Diagnostics & lifecycle

Where it sits in the software stack

Generic layered software of a central HPC (application processor with hypervisor, plus a safety island) and of a classic MCU-based ECU (AUTOSAR Classic). Highlighted modules matter for this topic. Click any module to open its page.

HPC · application processor (MPU) and safety island (MCU)
  • Type-1 hypervisor — Partitions CPU time, memory and devices between virtual machines and controls their communication.
  • BSP · drivers — VirtIO back-ends and physical device drivers support controlled device sharing between guest operating systems.
  • Cortex-A cores · GPU/NPU · LPDDR — Virtualization extensions and SMMU enforce guest execution boundaries and restrict device DMA access.
  • Own cores — Dedicated safety island cores run independently of guest virtual machines and hypervisor scheduling.
MCU-based ECU · AUTOSAR Classic

Not directly involved on this node type.