Arm Neoverse V3AE, with its exceptional computing performance, scalability, and security, empowers the next generation of vehicles and robots driven by NVIDIA DRIVE AGX Thor and Jetson Thor.
Smart devices across various industries are rapidly evolving. Vehicles are transforming into AI-defined platforms that support autonomous driving, AI-driven infotainment features, and personalized in-car experiences. Meanwhile, robotics is entering a new chapter—humanoid robots, autonomous mobile robots (AMRs), and intelligent industrial systems will all possess real-time perception, reasoning, and decision-making capabilities.
To realize the potential of smart devices, a “safety-first” computing platform is required, capable of delivering data center-level performance, scalability, and energy efficiency. With Arm Neoverse as the CPU core foundation, NVIDIA has launched a cutting-edge computing platform powered by NVIDIA DRIVE AGX Thor and Jetson Thor.
NVIDIA DRIVE AGX Thor: Driving Innovation in the Automotive Industry
The NVIDIA DRIVE AGX Thor developer kit is now available for purchase and will ship in September. Built on the Arm Neoverse V3AE CPU, NVIDIA DRIVE AGX Thor is a scalable and energy-efficient AI computing platform designed to handle the complex workloads required for autonomous driving.
The Arm computing platform provides NVIDIA DRIVE AGX Thor with deterministic performance and ISO 26262-compliant functional safety guarantees, ensuring that advanced AI workloads can run reliably in real-time. NVIDIA DRIVE AGX Thor achieves peak performance of up to 2,000 TFLOPS at FP4 precision, capable of simultaneously executing predictive advanced driver assistance systems (ADAS), real-time driving monitoring, natural voice interaction, and more on a single platform.
As the computational cornerstone of the NVIDIA DRIVE AGX Thor architecture, the Arm Neoverse V3AE CPU works closely with the NVIDIA Blackwell GPU and dedicated accelerators, laying a reliable and scalable foundation for the era of “AI-defined vehicles,” allowing cars to continuously learn, adapt, and optimize during use.
The NVIDIA DRIVE AGX Thor developer kit offers two versions: a desktop-level device (SKU10) and an automotive-grade device (SKU12), both of which accelerate the hardware and software development process.
NVIDIA Jetson Thor: Leading the Evolution of Robotics Technology
On the other hand, the launch of NVIDIA Jetson Thor extends the foundational advantages of Arm Neoverse V3AE into the realms of robotics and physical AI. NVIDIA Jetson Thor is designed for real-time, multi-modal AI workflows, enabling robots to simultaneously achieve perception, reasoning, and action—core requirements for humanoid robots, autonomous mobile robots, and industrial systems.
Arm Neoverse provides robotics technology with exceptional single-core performance and performance efficiency per watt, ensuring that humanoid robots and autonomous mobile robots can run multiple AI models in parallel without compromising response speed.
Examples of practical application scenarios:
Humanoid Robots: NVIDIA Jetson Thor provides the high-performance computing, low latency, and real-time reasoning capabilities required for humanoid robots, enabling them to run multiple models for vision, language, and motion planning in parallel, thus achieving natural interaction.
Industrial Automation: In factory and warehouse scenarios, NVIDIA Jetson Thor powers autonomous devices and robotic arms, enhancing efficiency and safety.
Medical Systems: With local sensor fusion technology and advanced AI capabilities, NVIDIA Jetson Thor supports medical robots in performing patient monitoring, rehabilitation training, and surgical assistance tasks.
Additionally, when used in conjunction with NVIDIA Blackwell, NVIDIA Jetson Thor achieves performance of up to 2,000 TFLOPS at FP4 precision, with computational performance increased by 7.5 times and energy efficiency improved by 3.5 times compared to NVIDIA Jetson Orin. With these core capabilities, NVIDIA Jetson Thor becomes a universal platform for deploying physical AI across various industries.
The NVIDIA DRIVE AGX Thor developer kit and NVIDIA Jetson Thor not only feature a CPU with performance increased by 3.1 times but also integrate more cores and stronger GPU performance than previous generations—these advantages provide robust support for real-time control in advanced robotics.
The Core Role of Arm Neoverse V3AE
The NVIDIA DRIVE AGX Thor developer kit and NVIDIA Jetson Thor exemplify how Arm Neoverse V3AE empowers data center-level performance for automotive and robotics workloads while balancing energy efficiency and safety at the chip level. Arm Neoverse V3AE boasts exceptional single-thread performance and scalability, capable of running multiple complex AI models in parallel. For safety-critical real-time tasks, Neoverse V3AE is designed to meet the strictest automotive-grade standards, providing determinism, isolation, and reliability for safety-critical environments.
With Arm’s leading performance per watt, these systems can sustainably run advanced AI workloads, whether deployed in vehicles or edge robots. This integrated advantage allows devices to fuse multi-sensor data, run generative AI, and make safe, real-time decisions in environments where response speed is critical.
Scalable Ecosystem for Developers
The scalability of any computing platform largely depends on its ecosystem. Both DRIVE AGX Thor and Jetson Thor rely on NVIDIA’s advanced software stack to achieve seamless operation on Arm’s efficient computing architecture.
NVIDIA DriveOS 7.0.3 for the automotive sector: now ISO 26262 certified, supports secure boot functionality, and provides TensorRT optimization for AI models and generative models.
NVIDIA Isaac for robotics, Metropolis for visual AI, and Holoscan for real-time sensor processing collectively facilitate the rapid development of intelligent robotic systems.
Combined with Arm’s powerful development tools and ecosystem support, developers can reduce development complexity, accelerate time-to-market, and focus on innovation rather than system integration.
Injecting Strong Momentum into Smart Devices Across Industries
The launch of the NVIDIA DRIVE AGX Thor developer kit and NVIDIA Jetson AGX Thor marks a new phase for smart devices centered around Arm. Specifically, in the automotive sector, this means endowing “AI-defined vehicles” with real-time adaptive capabilities, thereby enhancing safety, efficiency, and user experience; in the robotics sector, it also provides intelligent support with safety-ready capabilities for humanoid robots, autonomous devices, and industrial systems.
Arm Neoverse V3AE creates the possibility for scalable, secure, and energy-efficient computing, ensuring that fleet and factory robots can continuously run advanced AI workloads at the edge.
Arm is collaborating with NVIDIA to lay a solid foundation for safe and scalable computing—this will provide computational support for every mile driven and every task executed.