In an exciting development for the field of robotics, Analog Devices, Inc. (ADI) has partnered with NVIDIA to adopt the powerful Jetson Thor platform, marking a significant step forward in humanoid robotics. This collaboration aims to enhance the physical intelligence and reasoning abilities of robots, bringing us closer to having humanoid robots perform complex tasks in real-world environments.

The Breakthrough of Jetson Thor

NVIDIA's Jetson Thor platform is a groundbreaking technological achievement that redefines the capabilities of robotics with server-class AI compute power in a mobile power envelope. Equipped with a robust NVIDIA Blackwell GPU, transformer engine, Multi-Instance GPU (MIG), a 14-core Arm Neoverse V3AE CPU, and up to 128 GB of LPDDR5X memory, the Jetson Thor delivers 2070 FP4 TFLOPS. These specifications provide the high-throughput I/O required for real-time, dense multimodal data processing, bridging the gap between perception and reasoning.

ADI's Edge Sensing and Control

Analog Devices brings its expertise in edge sensing, precision motion control, power integrity, and connectivity to the collaboration. ADI's contributions allow for high-fidelity sensing and the precise control necessary for contact-rich manipulation in humanoid robots. This combination ensures that robots can execute tasks with high accuracy, harnessing the advanced computing power of Jetson Thor effectively.

ADI's sensor suite includes multimodal tactile sensing, ToF depth sensors, high-accuracy IMUs, joint encoders, and multi-axis force/torque sensors. When integrated with Jetson Thor's compute capabilities, these sensors provide the physical intelligence required to tackle sophisticated industrial tasks.

Closing the Sim2Real Gap with Isaac Sim

ADI and NVIDIA's partnership extends into the virtual realm with the implementation of NVIDIA Isaac Sim. This simulation environment allows developers to create and test robotic systems in a virtual setting before deploying them in the real world. By embedding robotics foundation models within ADI's development stack and simulating interactions, ADI and NVIDIA aim to close the Sim2Real gap, ensuring that virtual models behave consistently with their physical counterparts.

The key to this integration lies in high-quality sensor models and parameterized device behaviors that enhance the accuracy of simulations, which ADI optimizes to match NVIDIA's hardware specifications. By perfecting digital twins and policy reinforcement training in the simulation environment, the partnership enables faster transition from concept to deployment for humanoids.

Expanding Reasoning and Physical Intelligence

The foundation models facilitate the integration of various sensory modalities to enable adaptive, real-time decision making. This leap in cognitive robotic abilities is supported by the MIG-based workload partitioning in Jetson Thor, which isolates GPU slices for different tasks such as locomotion, perception, and action planning, ensuring functional decomposition seamlessly.

The demand for humanoid robots is burgeoning across sectors such as logistics, agriculture, and healthcare, where precise and dexterous manipulation are crucial. ADI and NVIDIA's collaborative efforts will be pivotal in meeting this demand by reducing the time from prototype to production using integrated simulation and hardware solutions.

Future Prospects and Availability

The collaboration between ADI and NVIDIA marks the beginning of a promising era for humanoid robotics. As Jetson Thor becomes more widely available, developers can expect a rich ecosystem built around ADI's high-fidelity sensing and NVIDIA's powerful computing. To learn more about availability or to explore potential use cases, developers are encouraged to contact the ADI robotics team or visit NVIDIA's official page for the latest updates on Jetson Thor.

As the partnership unfolds, it will not only transform the landscape of humanoid robotics but also set new benchmarks for integrating AI with physical real-time reasoning, paving the way for the robots of the future.