NVIDIA has just unveiled its new Jetson Nano Computer. Like the popular RaspBerry Pi system, it’s basically a mini-PC. However, it is specifically designed for robotics.
Jetson Nano was developed especially for robotics
The Jetson Nano mini PC is able to take full advantage of CUDA-X, including AI, speech recognition, sensor integration and more. This is a breakthrough for deep learning and robotics projects.
Jetson Nano is small enough to hold the hand, but it uses the latest NVIDIA technologies to take the industry to the next level. The Mini-PC supports more sensors and high-resolution inputs. From entertaining home projects for programmers to large-scale applications, anything is possible with this device. NVIDIA hopes it will be at the heart of its CUDA-X projects, drones, robots, and even toys.
Isaac SDK
The simulation suite known as Isaac SDK provides the brain of the Jetson Nano Robot with an ideal environment for a robot. In this environment you can learn to interact and train according to your needs. Based on deep learning, the robots learn and perfect themselves, the possibilities in this area are endless.
Visit our guide to the best vacuum cleaner robots on the market.
Jetson Nano features a Maxwell-based 128-core graphics processor and an ARM A57 quad-core processor that can deliver about 472 gigaflops of computing power for neural networks, high-resolution sensors, and other robotic functions while consuming 5W.
The kit can run Linux right out of the box and supports a variety of AI environments (including NVIDIA’s own, of course). It is equipped with 4 GB RAM, Gigabit Ethernet and I/O to add cameras or the like.
As expected, the price is the fundamental basis. Although the Nano is not as powerful as the high-end Jetson models, it is available for $99 for home users and $129 for business units. This is in the range of amateurs, inventors and students who are willing to spend a little more (but only a little more) on developing their own automated devices.