CES 2025: Arm teams with Nvidia to boost software-defined vehicles

A year ago, a keynote presentation quipped that the CES event had become a de facto motor show, and the 2025 edition has shown the acceleration of connectivity and processing into the automotive sector and autonomous driving in general – and to address the needs of both drivers and car manufacturers, processor firm Arm has furthered its collaboration with ecosystem companies in a partnership with leading artificial intelligence (AI) technology firm Nvidia.

Arm said that when it comes to offerings for the automotive market, generative AI (GenAI) applications are becoming paramount, and drivers are demanding increasingly advanced autonomous features and user experiences. This means automakers need simpler, more consolidated systems with server-class performance to handle the complexities of AI.

The first fruits of the collaboration will be the launch of the Nvidia Drive AGX Thor centralised compute system, built on Nvidia Blackwell architecture and the first service to use Arm Neoverse V3AE, the initial Neoverse central processing unit enhanced for automotive applications.

Arm Neoverse V3AE was launched in March 2024 as part of a series of automotive technologies designed to accelerate the development of AI-enabled vehicles by up to two years in an automotive industry the company said was undergoing seismic change and transformation.

Vehicles now and in the future, it said, are being defined by the electronic systems powering them, as they essentially become “a computer on wheels” and the most complex technology devices people are likely to own. Arm sees this complexity as being driven by the proliferation of AI and an exponential growth of software characterising software-defined vehicles (SDVs), requiring new levels of performance, efficiency, safety and security.

Slated to be available in production vehicles later in 2025, and combining Nvidia’s AI expertise with what Arm described as “the power and advanced capabilities” of its compute platform, Nvidia Drive AGX Thor is designed to deliver the AI capabilities that next-generation consumer and commercial fleets demand. Arm noted that automotive original equipment manufacturers – including BYD, Hyper, XPENG, Plus, Nuro, Waabi and WeRide – plan to adopt Nvidia Drive AGX Thor to power the next generation of SDVs.

Arm said that as algorithms evolve from traditional convolutional neural networks towards transformers and GenAI, Nvidia Drive AGX Thor enables fewer discrete stages of computation and delivers higher fidelity or faster response time for a given workload. Automotive GenAI applications enabled by Nvidia Drive AGX Thor include automated driving and parking, driver and passenger monitoring, and feature-rich cockpit capabilities, all on a centralised, safe and secure platform.

Moreover, Arm assured that when deployed in level-three or higher autonomous driving use cases, a single Nvidia Drive AGX Thor will deliver the same functionality as multiple leading-edge devices in vehicles today, leading to significant performance benefits for automakers.

“Every year at CES, we see many promising innovations aimed at tackling some of the greatest challenges across all technology markets, but what you might not know is that a huge number of them are built on the Arm compute platform – and the automotive industry is no exception … 94% of global automakers use Arm technology in their newest vehicle models,” said Dipti Vachani, senior vice-president and general manager of the automotive line of business at Arm.

“By continuing our longstanding collaborations with companies like Nvidia, we can deliver new levels of performance, efficiency and AI capabilities, making it easier for automakers to develop the next generation of cutting-edge SDVs.”