NVIDIA Halos OS is a safety-rated operating system for physical AI, forming the software layer of the NVIDIA Halos stack announced on June 22, 2026. Halos OS runs safety-critical tasks on the NVIDIA IGX Thor compute module with a built-in Functional Safety Island (FSI) rated to IEC 61508 SIL 3.
Halos OS is available in two configurations: pure Linux and a Linux+QNX hybrid. The Linux+QNX variant uses the deterministic QNX Neutrino RTOS kernel for hard real-time tasks, while Linux handles the compute-intensive AI layer. The architecture allows safety-critical software to run on an isolated hardware subsystem (FSI), independent of the main GPU compute stack.
Halos OS is designed for certification compliance with functional safety standards: IEC 61508 (SIL 3), ISO 13849 and ISO/IEC TR 5469 (AI in safety systems). TÜV Rheinland is inspecting the IGX Thor module and Halos OS for certification readiness — the results will be the first real-world test of the system's credibility. Compliance verification is handled by the ANAB-accredited (ISO/IEC 17020) NVIDIA Halos AI Systems Inspection Lab.
NVIDIA Halos Core on IGX is available in early access for registered developers in Linux and Linux+QNX configurations. The first customer is Agility Robotics — the fifth-generation Digit humanoid, targeted for late 2026, will achieve cooperative safety (operating alongside humans without physical barriers) via the IGX Thor module running Halos Core with certified software.
General-purpose or robotics-specialized operating system — Ubuntu, Ubuntu Core, ROS-aware distributions. Provides the platform for middleware, runtime, and robotic applications.
Real-time operating system with a deterministic scheduler and guaranteed interrupt latency. Typically used in motion control, safety-critical systems, ADAS, and industrial robotics (VxWorks, QNX Neutrino, RT-Linux, Zephyr).
A Runtime is the environment or execution layer used to run code, load libraries, manage dependencies, and operate applications or services — either in real time or during normal system operation. In robotics this includes real-time operating system (RTOS) runtimes, ROS 2 executor runtimes, containerised execution environments (Docker, podman), and embedded C++ runtimes on microcontrollers.
The functional safety role describes software responsible for ensuring that a robot operating alongside humans does not cause harm even in the event of hardware failure, software error or unforeseen environmental event. The safety component implements: deterministic execution of critical tasks (hard timing determinism), fault containment, continuous system integrity monitoring, transitions to safe failure states, and compliance with functional safety standards — IEC 61508 (SIL), ISO 13849 (Performance Level), ISO/IEC TR 5469 (AI in safety systems). It typically runs on an isolated hardware subsystem (e.g. an FSI island) independent of the main compute stack.
The real-time control role describes software responsible for closed control loops executed at a guaranteed, deterministic frequency — typically 1 to 10 kHz for current/torque loops in actuators. Unlike soft real-time, in a hard real-time regime missing a deadline is treated as a system error. Implemented on deterministic kernels (Linux PREEMPT_RT, QNX Neutrino RTOS) or dedicated microcontrollers. Critical for biped balance stability, force control, impedance control and functional safety.
Robot Control denotes the role of software responsible for motion control, command execution, coordination of actuating elements and the direct operational logic of the robot.
Device Integration denotes the role of software responsible for communication, configuration, initialization and handling of specific devices, sensors, controllers or hardware components within a robotic system.
The NVIDIA Halos family comprises a full functional safety stack for physical AI: the Halos OS operating system, Halos Core SDK, the hardware FSI on IGX Thor and an accredited inspection laboratory.
Agility Robotics — fifth-generation Digit humanoid (cooperative safety deployment targeted for late 2026, eventually fenceless operation at Amazon, GXO, Schaeffler, Toyota facilities).
Early access for registered developers. The Halos ecosystem includes over 40 companies (Infineon, Texas Instruments, TÜV Rheinland, UL Solutions).
QNX Neutrino — the BlackBerry QNX microkernel RTOS with a POSIX API. Dominant in automotive ADAS / infotainment (235M+ vehicles) and safety-critical robotics.
Halos OS requires the NVIDIA IGX Thor module with a Functional Safety Island (FSI). Does not run on standard Jetson modules without FSI.
License family: Proprietary – Commercial
No version history.
