Nvidia (NVDA.O) has launched a new compact generative AI supercomputer, the Jetson Orin Nano Super developer kit, which offers more performance at a lower price thanks to software upgrades.
This new product is small in size, only the size of the palm of your hand, and is suitable for a wide range of people from commercial AI developers to enthusiasts and students, improving the capabilities and performance of generative AI. Moreover, the price is only US$249 (about 1,800 yuan), a significant drop from the previous US$499.
This developer kit delivers a 1.7x improvement in generative AI inference performance, a 70% performance improvement to 67 INT8 TOPS, and a 50% increase in memory bandwidth to 102GB/s, a significant improvement over its predecessor.
Whether creating retrieval-augmented generative LLM chatbots, building visual AI agents, or deploying AI-based robots, Jetson Orin Nano Super is the ideal solution.
For users who already have the Jetson Orin Nano developer kit, the available software updates for the new Jetson Orin Nano Super will also enhance generative AI performance.
Jetson Orin Nano Super is ideal for those interested in developing skills in generative AI, robotics, or computer vision. As the AI field shifts from task-specific models to foundational models, it also provides an easy-to-use platform to turn ideas into reality.
Additionally, the enhanced performance of Jetson Orin Nano Super provides a boost for all popular generative AI models and transformer-based computer vision.
The developer kit includes an 8GB memory Jetson Orin Nano system module (SoM) and a reference carrier board, providing an ideal platform for edge AI application prototyping.
This SoM features NVIDIA's Ampere architecture GPU, equipped with tensor cores and a 6-core Arm CPU, supporting multiple parallel AI application pipelines and high-performance inference. It supports up to four cameras, offering higher resolution and frame rates than previous versions.
Generative AI is rapidly evolving. NVIDIA's Jetson AI Lab provides instant support for cutting-edge models from the open-source community and offers easy-to-use tutorials. Developers can also receive support from a wide Jetson community and gain inspiration from projects created by other developers.
The Jetson platform runs NVIDIA AI software including NVIDIA Isaac (for robotics), NVIDIA Metropolis (for visual AI), and NVIDIA Holoscan (for sensor processing). Development time can be shortened using NVIDIA Omniverse Replicator (for synthetic data generation) and the NVIDIA TAO Toolkit (for fine-tuning pre-trained AI models from the NGC catalog).
Partners in the Jetson ecosystem provide additional AI and system software, development tools, and custom software development services. They can also supply cameras and other sensors, as well as design services for carrier boards and product solutions.
To enhance generative AI performance by 1.7 times, software updates will also be released for the Jetson Orin Nano and Orin NX series system modules. Existing Jetson Orin Nano developer kit users can immediately upgrade the JetPack SDK to unlock the enhanced performance.
Article forwarded from: Jin Ten Data
