In 2019, as artificial intelligence technology rapidly advanced and AI applications became increasingly widespread, NVIDIA launched an AI computer and development kit called “Jetson Nano” aimed at “embedded designers, researchers, and DIY makers,” priced at $499.
▲ Jetson Nano (Source: NVIDIA)
Recently, NVIDIA released an upgraded version of the Jetson Nano – the “Jetson Orin Nano Super,” which is currently the “most affordable” AI supercomputer and development kit in the series. While offering improved performance, its price has been halved to just $249.
According to NVIDIA’s official statement, the Nano Super targets “commercial AI developers, hobbyists, and students.” Indeed, the price point of under $2,000 is quite accessible for those who are just getting started with generative AI, robotics, or computer vision technologies.
This perhaps indicates that the cost of performing AI inference tasks is gradually decreasing.
▲ Jetson Orin Nano Super (Source: NVIDIA)
Compared to its predecessor, the Nano Super offers 1.7x better AI inference performance and 70% higher overall performance, reaching 67 INT8 TOPS. Additionally, the Nano Super’s memory bandwidth has increased to 102GB/s, a 50% improvement over the previous generation. The CPU frequency has also been boosted from 1.5GHz to 1.7GHz.
The complete Nano Super development kit includes a Jetson Orin Nano 8GB system-on-module (SoM) and a reference carrier board. The SoM is equipped with an NVIDIA Ampere architecture GPU and a 6-core Arm CPU, supporting multiple concurrent AI application pipelines and high-performance inference.
Furthermore, it supports up to four cameras and provides higher resolution and frame rates compared to previous versions.
▲ Source: NVIDIA
NVIDIA claims that the Nano Super’s performance improvements benefit all popular generative AI models and transformer-based computer vision applications.
Moreover, since the Nano Super kit’s hardware is essentially identical to its predecessor, the Orin Nano, existing Orin Nano users can experience the Nano Super’s performance upgrades by installing a software update (JetPack SDK).
NVIDIA calls this a “gift that keeps on giving.”
▲ Source: NVIDIA
Within NVIDIA’s ecosystem, developers can access tutorials from NVIDIA Jetson AI Labs, get support from the broader Jetson community, and find inspiration from projects created by other developers.
NVIDIA states that whether creating retrieval-augmented generation-based LLM chatbots, building visual AI agents, or deploying AI-based robots, the Nano Super is the “ideal solution.” They expect it to accelerate the development of AI-driven robotics across multiple industries.
As the artificial intelligence field transitions from task-specific models to more general foundation models, the Nano Super undoubtedly provides technology enthusiasts with the “most affordable” accessible platform to turn their ideas into reality. As NVIDIA puts it:
“Now everyone can unlock new possibilities through generative AI.”