DayFR Euro

Nvidia has cooked a super-powerful “Raspberry Pi”: what is it for?

What if we ran ChatGPT or Midjourney on a small PC? This is the objective of Nvidia and its mini-supercomputer specialized in generative AI. All for a very low price of $249.

In a video posted on YouTube, Nvidia President Jensen Huang literally took the Jetson Orin Nano Super Developer Kit out of his oven, namely a development kit, or rather a small computer. A high-performance and very compact PC that allows generative AI to operate locally. Its goal: to integrate it into lots of connected objects.

Ce « Raspberry Pi » is able to run ChatGPT (or almost)

102 GB/s of bandwidth and 67 trillion operations per second: this is the performance of what looks like a Raspberry Pi. A motherboard, a fan, chips and ports: this is not the one of those small PCs, but indeed an Nvidia development kit.

The company sells it to commercial AI developers who would like to integrate it into their electronic devices. For the most knowledgeable, know that this development kit consists of an 8 GB system module (SoM), equipped with an Nvidia Ampere GPU and a 6-core Arm CPU which goes up to 1.7 GHz. An architecture that allows multiple AI applications to run simultaneously. For computer vision uses, the kit can support four cameras.

This content is blocked because you have not accepted cookies and other trackers. This content is provided by YouTube.
To be able to view it, you must accept the use carried out by YouTube with your data which may be used for the following purposes: allowing you to view and share content with social media, promoting the development and improvement of products from Humanoid and its partners, display personalized advertisements to you in relation to your profile and activity, define a personalized advertising profile, measure the performance of advertisements and content on this site and measure the audience of this site (find out more more)

By clicking on “I accept all”, you consent to the aforementioned purposes for all cookies and other trackers placed by Humanoid and its partners.

You can withdraw your consent at any time. For more information, we invite you to read our Cookies Policy.

I accept everything

Manage my choices

What Nvidia is proposing is to make generative artificial intelligence work “ locally » on this PC that fits in your hand. All on a small power supply: the device consumes 25 W, which remains reasonable. That is to say, no Internet connection required, without using remote servers. Whether chatbots powered by LLMs, image generators or others, Nvidia presents its Jetson Orin Nano Super as the “ ideal solution ».

The power of the latter does not allow it to operate a “ TRUE » ChatGPT or a « TRUE » Midjourney, but slightly smaller models.

The Jetson Orin Nano Super from Nvidia // Source : Nvidia

We can very well imagine this card present in a few years inside various devices. A kitchen robot to create recipes with the food you have in your fridge, for example. Camera management could make it possible to have an intelligent video surveillance system, but whose data is not posted online on a service. The head of Nvidia even imagines fully autonomous robots, running on this small computer.

Why Nvidia sells this development kit

The Jetson Orin Nano Super is the perfect support for the company to sell its software: Isaac for robotics, Metropolis for visual AI, Holoscan for camera processing, or even LATTE3D for 3D generation of objects.

Jensen Huang explaining what this little PC is to his dogs // Source: Nvidia on YouTube

Enough to create a software and hardware ecosystem for Nvidia, which has entered the race to become an artificial intelligence giant. In 2024, the company experienced a meteoric rise in power: its stock rose 160% in one year. In terms of market capitalization, the company sits at $3.19 billion, behind Apple and Microsoft. Good financial results due to the growing demand for GPUs, which are used to run generative AI in data centers. With this small PC, the company also wants to conquer electronic devices which will become even more complex. Clearly, Nvidia has all the weapons to confirm its lead.


-

Related News :