NVIDIA DriveOS SDK
NVIDIA DriveOS™ is an automotive operating system developed with industry-standard safety and security methodologies certified by the globally renowned automotive certification organization, TÜV SÜD. This powerful platform is designed to meet cutting-edge automotive product requirements, including advanced AI inference, high-performance computer vision, advanced graphics, high-end audio, and complex safety and security use cases. DriveOS also supports autonomous driving and AI-powered cockpit experiences.
DriveOS enables Linux or QNX as the application operating system with a software development kit for image processing, diverse sensor integration, AI acceleration, low-overhead interprocess communication, and developer tools for debugging and profiling.
Requires NVIDIA DRIVE AGX™ SDK Developer Program membership
Benefits
Programmability
Enables a smooth transition from cloud or workstation to the SoC
Supports NVIDIA® CUDA® and TensorRT®, widely adopted programming models with a large developer base.
Supports unified APIs, from cloud to target.
Safety and Reliability
Designed for safety-critical applications
Complies with ASPICE, ISO 26262, and ISO/SAE 21434.
Supports heterogeneous redundancy for compute workloads.
Highly Optimized
Efficient processing of time-critical workloads
Camera frames are directly loaded into GPU memory for high-performance sensor interfacing and processing with NvMedia.
Supports NvStreams for efficient data transport, with zero-copy data transfer between hardware accelerators.
Enables highly flexible sensor-processing pipelines.
Accelerated Application Development
Building blocks and tools for automotive software
DriveWorks provides optimized algorithms and tools.
NVIDIA Nsight provides tools for debugging, profiling, and tracing.
Offers rich documentation including numerous samples to demonstrate features, webinars, and GTC talks to get familiar with DriveOS.
Virtualization and Containerization
Comprehensive scalability and isolation features
Hypervisor manages resources and provides abstraction between underlying hardware and OS on top.
Guest operating systems provide additional isolation and redundancy.
Host and target side Docker container support enable efficient development.