Nvidia has released a new $249 version of its Jetson computer designed to run synthetic applications locally. The palm-sized Orin Nano is said to double the speed and efficiency of its predecessor at half the price, and can handle nearly 70% more computational tasks, according to Nvidia.
The Orin Nano is ideal for hobbyists looking to train their own AI applications, or for developers of robotics and other industrial tools to run cutting-edge applications without connecting to the cloud.
short video On YouTube announcing the product, Nvidia CEO Jensen Huang stands in his kitchen pulling a tray out of his oven to reveal the palm-sized mini-computer. Huang continues, saying that the computer can process approximately “seventy trillion” operations per second and consumes only 25 watts of power.

“A long time ago, we created a completely new type of processor, which was an automated processor,” Huang says. “No one understood what we were building at the time, and we imagined that one day these deep learning models would evolve and we would have robots for everything.” Jetpack computers were intended to run robots, but today they can also run large language models like Meta’s Llama.
Sam Altman’s startup World (formerly known as Worldcoin), which aims to verify human identity through iris scans, It uses the Jetpack module Inside his orb scanning device. In a Blog post Back in October, World said: “In its latest release, Orb is equipped with the most advanced NVIDIA Jetson module with nearly 5 times the AI performance compared to the previous version to enable faster and smoother proof of human verification processes.”
In essence, the Orin Nano is a portable brain that can be connected to other devices to power its own AI functions. Cloud scalers like Amazon and Google charge fees for access to servers and AI models, and these costs can add up. Some applications may need guaranteed uptime and minimal latency, such as warehouse robots. Connecting to a cloud hosting provider remotely is not ideal. However, a computer like the Orin Nano will only be able to run more lightweight AI applications; it will not replace high-end Nvidia GPUs that cost tens of thousands of dollars and are capable of training and inferring large-scale AI models. .
However, if you are the type of person who wants to create some kind of new AI-powered robot, the Orin Nano might be for you.
https://gizmodo.com/app/uploads/2024/11/Jensen-Huang-Nvidia-1.jpg
Source link