Nvidia GTC: New Jetson module offers 200 TOPS for embedded edge AI - Embedded.com

Nvidia GTC: New Jetson module offers 200 TOPS for embedded edge AI

New Jetson AGX Orin module built on Nvidia Ampere architecture claims to be smallest and most powerful AI supercomputer for embedded edge computing.

Announced at Nvidia GTC is the Nvidia Jetson AGX Orin, which the company said is the world’s smallest, most powerful and energy-efficient AI supercomputer for robotics, autonomous machines, medical devices and other forms of embedded computing at the edge.

Built on the Nvidia Ampere architecture, Jetson AGX Orin provides 6x the processing power and maintains form factor and pin compatibility with its predecessor, Jetson AGX Xavier. It delivers 200 trillion operations per second, similar to that of a GPU-enabled server but in a size that fits in the palm of your hand.

The new Jetson computer accelerates the full Nvidia AI software stack, allowing developers to deploy the largest, most complex models needed to solve edge AI and robotics challenges in natural language understanding, 3D perception, and multisensor fusion.

“As robotics and embedded computing transform manufacturing, healthcare, retail, transportation, smart cities and other essential sectors of the economy, the demand for processing continues to surge,” said Deepu Talla, vice president and general manager of embedded and edge computing at Nvidia. “Jetson AGX Orin addresses this need, enabling the 850,000 Jetson developers and over 6,000 companies building commercial products on it to create and deploy autonomous machines and edge AI applications that once seemed impossible.”

Nvidia Jetson AGX Orin Module and Developer Kit 1

Jetson AGX Orin features an Nvidia Ampere architecture GPU and Arm Cortex-A78AE CPUs, along with next-generation deep learning and vision accelerators. High-speed interfaces, faster memory bandwidth and multimodal sensor support provide the ability to feed multiple concurrent AI application pipelines.

Software and ecosystem support
Customers using the Jetson AGX Orin can leverage the CUDA-X accelerated computing stack, JetPack SDK and the latest Nvidia tools for application development and optimization, including cloud-native development workflows. Pre-trained models from the Nvidia NGC catalog are optimized and ready for fine-tuning with the TAO toolkit and customer datasets. This reduces time and cost for production-quality AI deployments, while cloud-native technologies allow seamless updates throughout a product’s lifetime.

The Jetson embedded computing partner ecosystem encompasses a broad range of services and products, including cameras and other multimodal sensors, carrier boards, hardware design services, AI and system software, developer tools and custom software development.

“Driven by the advancements in AI, the robotics market is growing exponentially,” said Jim McGregor, principal analyst at TIRIAS Research. “Nvidia is the recognized leader in AI and continues to leverage this expertise to advance robotics through a robust ecosystem and complete end-to-end solutions, including a range of hardware platforms that leverage common tools and neural network models. The new Jetson platform brings the performance and versatility of the Nvidia Ampere architecture to enable even further advancements in autonomous mobile robots for a wide range of applications ranging from agriculture and manufacturing to healthcare and smart cities.”

For specific use cases, software frameworks include Nvidia Isaac Sim on Omniverse for robotics, Nvidia Clara Holoscan SDK for healthcare, and Nvidia DRIVE for autonomous driving. The latest Isaac release includes significant support for the Robot Operating System (ROS) developer community. Nvidia has also released the new Omniverse Replicator for synthetic data generation and Isaac GEMs, hardware-accelerated software packages that make it easier for ROS developers to build high-performance AI-enabled robots on the Jetson platform.

Real-time sensing for medical instruments, devices
The AGX Orin powers Nvidia Clara Holoscan, a new computing platform for the healthcare industry that allows developers to build software-defined medical devices which run low-latency streaming applications on the edge. This processing power is needed for devices such as robotic surgery, endoscopy and diagnostic imaging systems to process physics, imaging and visualization and enable real-time AI-decision support for doctors. Clara Holoscan gives developers a flexible platform to create AI microservices that run low-latency streaming applications on devices while passing more complex tasks to the data center.

Transportation AI assistants
The DRIVE AGX Orin, also powered by the Ampere architecture like Jetson AGX Orin, is a platform for the transportation industry. It is the advanced processor behind the newly announced Nvidia DRIVE concierge and DRIVE chauffeur — two AI platforms dedicated to redefining the passenger experience inside the car through AI assistants, and to powering safe, autonomous driving, respectively. Numerous global auto and truck makers, new energy vehicle startups and shuttle companies are using its high-performance AI compute for their next generation of intelligent, software-defined mobility solutions.

The Jetson AGX Orin module and developer kit will be available in the first quarter of 2022. 

Related Contents:

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.