NVIDIA rolls out DRIVE AGX Thor developer kit for automotive Artificial Intelligence development

NVIDIA has opened preorders for the DRIVE AGX Thor developer kit, a Blackwell-architecture platform purpose-built to accelerate automotive Artificial Intelligence development, with delivery expected in September. The kit targets developers and researchers building advanced sensor and reasoning models for autonomous vehicles.

NVIDIA has made the DRIVE AGX Thor developer kit available for preorder, with delivery slated for September. The kit is presented as a development and research platform to accelerate design, testing and deployment of autonomous vehicle systems that increasingly rely on reasoning vision language action models, generative Artificial Intelligence and advanced sensors. NVIDIA positions the kit as automotive ready, offering sufficient I/O for surround cameras, radars and lidars and standard vehicle interfaces such as GbE/10GbE and PCI-Express.

The developer kit is built on NVIDIA s Blackwell architecture, Arm Neoverse V3AE central processing units and the NVIDIA DriveOS 7 software stack. NVIDIA says DRIVE AGX Thor is purpose built for reasoning vision language action models and meets automotive functional safety and cybersecurity standards, specifically ISO 26262 for functional safety and ISO 21434 for cybersecurity. NVIDIA also links the platform to the broader DRIVE AGX family and to other developer offerings, including Jetson AGX Thor developer kits and Jetson T5000 modules for robotics and physical Artificial Intelligence work.

DRIVE AGX Thor already has a broad ecosystem of automakers, autonomous trucking firms, suppliers and technology partners building on the platform. Named customers and partners in the announcement include BYD, GAC, IM Motors, Li Auto, Volvo Cars, Xiaomi and Zeekr, and trucking or logistics companies such as Aurora, Gatik, PlusAI and Waabi. Tier 1 and production system suppliers listed include Continental Automotive, Desay SV, Lenovo, Magna and Quanta, and the platform is supported by sensor and embedded technology vendors such as AdaCore, Lauterbach, OMNIVISION, QNX and Vector. NVIDIA ties the kit to its Halos safety system, which combines platform, algorithmic and ecosystem safety and uses NVIDIA DGX for training, NVIDIA Omniverse and NVIDIA Cosmos on NVIDIA OVX for simulation, and DRIVE AGX for deployment. Developers are invited to watch the unboxing video and join the NVIDIA DRIVE AGX SDK Developer Program to get started.

73

Impact Score

Saudi Artificial Intelligence startup launches Arabic LLM

Misraj Artificial Intelligence unveiled Kawn, an Arabic large language model, at AWS re:Invent and launched Workforces, a platform for creating and managing Artificial Intelligence agents for enterprises and public institutions.

Introducing Mistral 3: open artificial intelligence models

Mistral 3 is a family of open, multimodal and multilingual Artificial Intelligence models that includes three Ministral edge models and a sparse Mistral Large 3 trained with 41B active and 675B total parameters, released under the Apache 2.0 license.

Contact Us

Got questions? Use the form to contact us.

Contact Form

Clicking next sends a verification code to your email. After verifying, you can enter your message.