Nvidia’s GTC conference put autonomous driving and industrial AI near the center of the company’s latest push. Across several announcements, automakers and self-driving vehicle companies described how they plan to use Nvidia products in vehicle production, factory simulation, robotics, driver-assistance systems and autonomous trucking.
The announcements show Nvidia trying to serve more than one layer of the transportation stack. Its tools are being used inside vehicles, in software testing, in factory planning and in engineering simulations.
GM brings Nvidia deeper into factories and vehicles
General Motors said it has expanded its partnership with Nvidia in a collaboration that reaches across factories, robots and self-driving cars. A major part of the plan is the use of Omniverse with Cosmos to train AI manufacturing models and support next-generation factory development.
For GM, Omniverse is aimed at digital factory work. The automaker plans to build digital twins of factories so it can virtually test new production processes without interrupting existing vehicle production. That matters because automakers need to change manufacturing systems while still keeping current vehicle lines moving.
GM also said it will use Omniverse to train robotics platforms for operations such as material handling and transport. In that context, Nvidia’s role is not only about cars on roads. It also extends to the machines and processes that help build those cars.
Inside future vehicles, GM said it will use Nvidia’s Drive AGX as in-vehicle hardware for future advanced driver-assistance systems and in-cab in safety experiences. That puts Nvidia’s technology into the production and user-facing sides of the automaker’s business.
Gatik targets autonomous middle-mile logistics
Gatik, a self-driving truck company backed by Isuzu and Goodyear Ventures, has joined Nvidia’s automotive ecosystem. The Silicon Valley and Toronto-based company focuses on autonomous middle-mile logistics using self-driving box trucks.
Gatik said it will develop and deploy Drive AGX, accelerated by Drive Thor, as the AI brain across its fleet of trucks. The company also said it is running its AI models on the DriveOS system for safety.
The collaboration is meant to help speed deployment of Level 4 autonomous trucks at scale for Gatik’s customers. The source names Walmart, Kroger and Tyson Foods among those customers.
The key point is specialization. Gatik is not presented as building a general consumer robotaxi service. Its work is tied to middle-mile logistics, where self-driving box trucks move goods through commercial routes.
Plus and Torc build toward autonomous trucking launches
Plus, an autonomous trucking software startup, said it will use Cosmos world foundation models to accelerate testing and development of SuperDrive, its autonomous driver. The company said SuperDrive is built on multiple Nvidia Drive Orin SoCs.
David Liu, co-founder and CEO of Plus, told TechCrunch: “Today we’re doing production deployments with Orin and we’re working with our suppliers on producing Thor systems [in the future].”
Plus has recently made deals with commercial vehicle manufacturers, including Traton Group, IVECO and Hyundai, to integrate SuperDrive into their trucks. The company has been testing its technology on public roads in Texas and Sweden and has targeted a 2027 commercial launch.
Torc, another self-driving truck company, also announced work with Nvidia. The Virginia-based company, a subsidiary of Daimler Truck AG, is developing a scalable physical AI compute system for its autonomous vehicles. Torc will also work with Flex, which builds automotive-grade compute platforms.
Torc said it is using Nvidia chip architecture and software including Drive AGX, Drive Orin and DriveOS. The goal is to support future deployment of autonomous driving capabilities as the company works toward a 2027 commercial launch. In October 2024, Torc achieved its first driver-out test on a closed course in Texas.
Volvo uses Nvidia GPUs for faster design simulation
Volvo’s announcement was different from the automated driving collaborations. The automaker is using Nvidia’s Blackwell GPUs to power aerodynamics simulations rather than to accelerate automated driving technology.
The source explains the practical reason: better aerodynamics, especially in an EV, can improve battery performance by reducing drag on the car. Faster simulation gives designers more room to test and refine vehicle shapes.
Volvo is working with Ansys, a software simulation company, rather than using Nvidia’s Omniverse simulator. Ansys’s “Fluent” simulation software runs on Nvidia’s Blackwell GPUs.
Together, Volvo and Ansys used a CAD model for the EX90 as a data input for Fluent simulations. That helped Volvo reduce total simulation run time from 24 hours to 6.5 hours, enabling multiple design iterations per day, optimized vehicle design and accelerated time to market.
Halos puts Nvidia’s safety tools under one umbrella
Nvidia also unveiled Halos, which it defines as an AI-powered safety system for autonomous vehicles and future physical AI, including humanoid robots. Halos brings together several of Nvidia’s automotive hardware and software safety solutions.
That positioning is important because the announcements span many environments: factories, trucks, passenger vehicles, engineering tools and future physical AI systems. Halos is presented as an umbrella for safety-related Nvidia tools across those uses.
Across GM, Gatik, Plus, Torc and Volvo, the pattern is clear from the source: Nvidia is not only selling chips for vehicles. It is offering a wider automotive platform that touches simulation, production, robotics, in-vehicle compute and autonomous driving development.