NVIDIA GTC 2025 Auto/Autonomous Roundup Halos, GM, Ansys, Volvo, Plus, Nexar, Gatik, Uber & Tor

At the NVIDIA GTC conference in San Jose, leading mobility companies showcased advancements powered by NVIDIA’s AI technologies across passenger vehicles, trucks, and autonomous systems. The company announced Halos Safety for AVs. Automakers, including General Motors and Volvo Cars, are leveraging NVIDIA’s platforms — DGX for AI model training, Omniverse and Cosmos for simulation, and DRIVE AGX for in-vehicle computing — to optimize manufacturing and enhance vehicle safety.

In the trucking sector, NVIDIA DRIVE AGX is enabling autonomous operations for companies like Gatik, Uber Freight, and Torc, while Magna is integrating DRIVE AGX Thor to meet growing demand for AI-driven vehicle platforms. NVIDIA’s Omniverse Blueprint, expanded with Cosmos world foundation models, is accelerating AV simulation and validation for companies like Plus, Foretellix, and Mcity.

AI-powered in-vehicle experiences are also advancing, with Cerence AI and SoundHound demonstrating next-generation voice assistants that leverage NVIDIA AI Enterprise and DRIVE AGX for cloud and edge-based generative AI capabilities.

NVIDIA Unveils Halos: Comprehensive Safety System for Autonomous Vehicles

NVIDIA has launched NVIDIA Halos, a full-stack safety system designed to ensure the safe development and deployment of autonomous vehicles (AVs). Halos integrates NVIDIA’s automotive hardware, software, AI models, and tools across all stages—from AI training in the cloud to in-vehicle deployment. The system focuses on three key areas: platform safety, algorithmic safety, and ecosystem safety, combining safety-assessed chips, AI-powered models, and validated workflows to create safer AV systems.

Halos operates on three levels:

  • Technology Level: Covers platform, algorithmic, and ecosystem safety.
  • Development Level: Provides guardrails during design, deployment, and validation.
  • Computational Level: Utilizes NVIDIA DGX for AI training, NVIDIA Omniverse and Cosmos for simulation, and NVIDIA DRIVE AGX for deployment.

An essential part of Halos is the AI Systems Inspection Lab, which verifies safe integration of AV systems with NVIDIA’s technology. Accredited by the ANSI National Accreditation Board, it includes inaugural members like Ficosa, OMNIVISION, onsemi, and Continental.

Halos leverages NVIDIA’s extensive safety expertise, including 15,000+ engineering years invested in vehicle safety, 10,000+ hours contributing to international standards, and 1,000+ AV safety patents. It also builds on recent safety certifications, such as ISO 26262 ASIL D and ISO/SAE 21434 cybersecurity standards. With Halos, NVIDIA is driving advancements in AV safety, helping partners accelerate standardization and regulatory compliance in the rapidly evolving autonomous transportation sector.

GM and NVIDIA Partner to Advance AI in Vehicles, Factories, and Robotics

General Motors (GM) and NVIDIA are collaborating to revolutionize vehicle manufacturing and autonomous technology using AI, simulation, and accelerated computing. The partnership will leverage NVIDIA’s platforms, including NVIDIA Omniverse™ and NVIDIA Cosmos™, to train AI models for optimizing GM’s factory planning and robotics. GM will also integrate NVIDIA DRIVE AGX™ for future advanced driver-assistance systems (ADAS) and in-cabin safety enhancements.

GM plans to use the NVIDIA Omniverse platform to create digital twins of assembly lines, allowing virtual testing and production simulations that improve efficiency and reduce downtime. Additionally, AI-powered robotics will enhance material handling, transport, and precision welding, boosting safety and operational performance. Future GM vehicles will incorporate NVIDIA DRIVE AGX, powered by the Blackwell architecture and running DriveOS™, capable of delivering 1,000 trillion operations per second to accelerate autonomous vehicle (AV) deployment.

Leaders Mary Barra (GM) and Jensen Huang (NVIDIA) highlighted that this partnership combines technology and human craftsmanship to transform transportation and manufacturing. The extended collaboration will be showcased at the NVIDIA GTC global AI conference, where the companies will discuss the impact of AI on automotive innovation.

General Motors (GM) is set to develop its next-generation vehicles using NVIDIA DRIVE AGX, powered by the advanced NVIDIA Blackwell architecture and running the safety-certified NVIDIA DriveOS operating system. This powerful in-vehicle computing platform is designed to handle the immense computational demands of autonomous vehicle (AV) systems, ensuring both safety and performance.

With the ability to process up to 1,000 trillion operations per second, NVIDIA DRIVE AGX significantly accelerates the development and deployment of safe AVs at scale. By leveraging this high-performance platform, GM can streamline its efforts to bring reliable autonomous technology to market more efficiently, enhancing safety and driving innovation in the automotive sector.

Ansys, Volvo Cars, and NVIDIA Achieve Breakthrough in Aerodynamic Simulations

Ansys (NASDAQ: ANSS), in collaboration with Volvo Cars and NVIDIA, has achieved a major breakthrough in aerodynamics simulations, reducing simulation run time from 24 hours to just 6.5 hours using eight NVIDIA Blackwell GPUs. This advancement allows Volvo Cars to conduct multiple design iterations per day, enhancing optimization studies for battery electric vehicles (BEVs) and accelerating time-to-market. High-fidelity CFD simulations, crucial for reducing aerodynamic drag and improving EV range, were previously time-consuming and costly. By scaling Ansys Fluent to eight GPUs, meshing time was reduced to one hour and solver time to 5.5 hours, delivering a 2.5x speed increase over traditional CPU-based hardware.

This accelerated process helps Volvo Cars optimize designs faster, meeting critical efficiency and emissions standards like WLTP. Ansys and NVIDIA leaders highlighted how GPU-accelerated simulations can drive innovation across industries, including automotive, aerospace, and consumer electronics, by enabling faster, high-precision simulations and improving overall product performance.

Nexar and NVIDIA Partner to Accelerate Autonomous Vehicle AI Development

Nexar, a leader in AI-powered mobility solutions, has partnered with NVIDIA to advance autonomous vehicle (AV) technology by leveraging real-world vision data and AI-powered simulations. Current physical vehicle testing is limited and costly, making it difficult to encounter rare, dangerous scenarios necessary for training AVs. Through AI and accelerated computing, semantic video search and simulation can generate useful datasets to improve AV safety and performance.

By combining Nexar’s extensive real-world data—capturing over 100 million miles monthly—with NVIDIA’s AI and simulation capabilities, the collaboration enhances AV training, high-definition mapping, and predictive modeling. This partnership also aims to strengthen NVIDIA Cosmos world foundation models and accelerate Nexar’s AI model training through NVIDIA DGX Cloud. Leaders from both companies emphasized that this collaboration will pave the way toward fully autonomous mobility by refining AV foundational models and improving safety across the industry.

Plus and NVIDIA Partner to Advance AI for Autonomous Trucks

Plus, a global leader in AI-based autonomous trucking software, is collaborating with NVIDIA to integrate NVIDIA Cosmos™ world foundation models (WFMs) into its SuperDrive™ virtual driver system, accelerating the development of Level 4 self-driving trucks. SuperDrive, built on the NVIDIA DRIVE AGX platform, combines generative AI and explainable systems to ensure high performance and safety. It has already been selected by leading manufacturers such as TRATON GROUP, IVECO, and Hyundai to power their autonomous vehicle programs.

As Plus advances toward launching factory-built driverless trucks by 2027, it will use NVIDIA Cosmos to enhance its models for handling novel driving scenarios, boosting the readiness of its in-vehicle technology for commercial deployment. This collaboration positions Plus to accelerate the validation of new commercial routes and ensure scalable, safe, and reliable autonomous trucking. Leaders from Plus and NVIDIA emphasized that this partnership will democratize physical AI and revolutionize the future of autonomous transportation.

NIVIDIA Keeps on Truckin’

For trucking NVIDIA DRIVE AGX, which provides the computing power needed for safe and reliable autonomous operations.

Companies like Gatik are using DRIVE AGX to power AI processing in their driverless delivery trucks, which serve major brands such as Tyson Foods and Kroger.

Uber Freight is also adopting this platform to enhance efficiency across its carrier fleets.

Torc is preparing to roll out a scalable AI system for autonomous trucks by 2027, leveraging NVIDIA’s technology.

The demand for NVIDIA DRIVE AGX is growing rapidly, with Magna leading the charge in integrating this technology into future vehicles.

Magna is developing driving systems using NVIDIA DRIVE Thor, which is based on the high-performance NVIDIA Blackwell architecture, enabling advanced functions like generative AI and language models for in-vehicle experiences.

Meanwhile, NVIDIA’s Omniverse Blueprint for AV simulation is expanding, with companies like Plus and Foretellix using it to refine their autonomous driving systems. The blueprint’s integration with CARLA, an open-source AV simulator, and Capgemini’s adoption of its capabilities signal a major leap forward in developing and validating AV technologies.