Jakarta, Gizmology – NVIDIA has reaffirmed its dominance in the world of artificial intelligence by introducing NVIDIA IGX Thor, a next-generation industrial AI platform designed to bring real-time physical AI capabilities directly to real-world environments. This platform marks a significant milestone in AI's shift from the digital realm to the physical world, from manufacturing plants to hospital operating rooms.
With up to eight times the power of its predecessor, the IGX Orin, the system is claimed to process high-speed sensor data, maintain industrial-grade reliability, and ensure functional safety in a compact module size. Several major companies such as Hitachi Rail, Joby Aviation, Diligent Robotics, and CMR Surgical have become early partners in adopting this technology, marking a real step towards a future of smarter human-machine collaboration.
NVIDIA IGX Thor is more than just a performance boost; it's a new foundation for an era where robotics, medical systems, and industrial automation require fast processing without compromising security. This platform is expected to accelerate the development of generative AI solutions in physical environments, from data-driven surgery to autonomous decision-making in autonomous vehicles.
Also Read: Qualcomm Announces Snapdragon 6s Gen 4, a Mid-Range Chipset for Better Performance
AI Performance Up to 8 Times Faster than Previous Generation
IGX Thor is designed to address challenges faced by the industrial and medical sectors, particularly the limitations of edge computing systems in securely processing real-time data. The platform delivers a significant performance leap thanks to the combination of two GPUs based on the NVIDIA Blackwell architecture: an integrated GPU (iGPU) and a discrete GPU (dGPU).
With a total of 5.581 FP4 TeraFLOPS and 400 GbE connectivity, NVIDIA IGX Thor delivers up to an 8x performance increase on iGPUs and 2,5x performance on dGPUs compared to IGX Orin. This speed enables seamless processing of large language models (LLMs) and vision-language models (VLMs) in industrial environments without relying entirely on cloud computing.
In addition to its performance, NVIDIA IGX Thor is built to high industry standards with a 10-year lifecycle and long-term support for the NVIDIA AI software stack. The platform runs the NVIDIA AI Enterprise stack, including NVIDIA NIM microservices, and integrates with the Isaac ecosystem for robotics, Metropolis for visual AI, and Holoscan for medical sensor processing.
Improving Safety and Efficiency in the Industrial and Medical Sectors
NVIDIA also includes the full NVIDIA Halos safety system, which enables IGX Thor to handle human-machine collaboration with a high level of security. This feature leverages data from various internal and external sensors to ensure safe human-machine interactions—a crucial element in environments like factories or operating rooms.
Several major companies have begun testing NVIDIA IGX Thor. Hitachi Rail is using the platform for predictive maintenance and automated inspections on its rail network, aiming to improve operational efficiency and reliability. Meanwhile, Maven Robotics plans to integrate IGX Thor into its next-generation industrial robots to deliver AI with environmental awareness and autonomous decision-making capabilities.
In the medical sector, CMR Surgical, Diligent Robotics, and EndoQuest Robotics are pioneers testing IGX Thor's capabilities to power their surgical robotic systems. With its real-time processing power, the platform enables complex data analysis during surgery, helping doctors make quick and accurate decisions without compromising patient safety.
Global Ecosystem Support for Real-World AI Adoption
NVIDIA isn't working alone in bringing IGX Thor to market. A number of global hardware manufacturers and solution providers, including Advantech, ADLINK, Curtiss-Wright, ASRock Rack, and NEXCOM, are expanding the ecosystem by offering edge servers, development boards, and camera and sensor modules optimized for the platform.
IGX Thor will be available in two production-ready systems: the NVIDIA IGX T5000 module and the NVIDIA IGX T7000 board kit. Both are designed for a wide range of AI physics applications with built-in functional safety, from industrial robots to precision medical systems. NVIDIA confirms availability of developer kits and production systems will begin in December 2025.
With the launch of NVIDIA IGX Thor, NVIDIA is solidifying its position as a key driver of the industrial and medical AI revolution. IGX Thor is more than just powerful computing hardware; it's the foundation of an AI ecosystem that enables companies to create autonomous solutions with greater safety, speed, and efficiency in the real world.
NVIDIA IGX Thor is a concrete step in bringing AI from the digital realm to the physical world. With significantly higher performance, comprehensive software support, and industry-standard safety, this platform paves the way for a new generation of intelligent machines, from factory robots to automated surgical systems.
For them, NVIDIA IGX Thor isn't just about artificial intelligence, but about delivering AI that can truly "feel" and "think" in the real world. A vision that heralds a new era, where human-machine collaboration is no longer a futuristic concept, but a reality being built today.
Explore other content from Gizmologi.id
Subscribe to get the latest posts by email.
