Nvidia, on Tuesday, unveiled its most affordable generative AI supercomputer called Jetson Orin Nano Super Developer Kit. Priced at $249 (approximately Rs 21,146), this tiny AI accelerator offers up to 67 trillion operations per second (TOPS) of AI performance, which is more than that of most AI PCs priced over $1,000.
It is currently available for students, enthusiasts, and developers across the world. This isn’t a new product, as the Jetson Orin Nano Super was first announced in 2023 for $499. Now, Nvidia, a global leader in AI hardware and software, has slashed its price to $249 and the new software update has improved its performance from 40 TOPs to 67 TOPS.
Advertisement
The product has already gone out of stock in many parts of the world, with shipping time ranging from anywhere between four to five weeks.
“It runs everything that the Hyperscale Graphics Extension (HGX) does. Even runs large language models. And, can’t wait for all of you to try it,” said Jensen Huang, Nvidia CEO, in a YouTube video.
Nvidia makes AI accelerators that power the world’s leading large language models from OpenAI, Meta, Google, and many more. However, these accelerators are quite expensive, where the Nvidia HGX H100 costs Rs 34,00,000 or more which most small-scale developers cannot afford.
Advertisement
The Nvidia Jetson Orin Nano Super comes with 8 GB of system-on-module (SoM) memory and the graphics chip is based on Ampere architecture paired with a 6 core Arm CPU, capable of offering AI inferencing. It is also compatible with all the NVIDIA AI software and can run various AI models, including advanced robotics and transformer models. It is primarily designed to power advanced robotics applications and is also compatible with Vision AI, capable of powering a variety of Internet of Things (IoT) devices.
Jetson Orin Nano Super does not come with built-in storage. A user needs to download the operating system to a microSD card to boot the system, similar to a Raspberry Pie.
Nvidia Jetson Orin Nano is tiny and affordable. (Image credit: Nvidia) 
Share
