Pandora is “mini AI hardware” built into a mini PC form factor It uses Nvidia’s Jetson Orin NX Super platform It comes with a 128GB SSD and Ubuntu OS in a less than 500g chassis
Palit, Nvidia’s largest add-in board partner, has been a leading supplier to the global video card market for more than 30 years. Best known for its Nvidia GeForce graphics card lineup, the company is now expanding into the AI space with Pandora, a compact mini-PC device built for edge AI computing.
Pandora (announced in a YouTube video spotted by Videocardz) measures just 121mm x 145mm x 66mm and weighs 470g. Although similar to a mini PC, the new device is called “mini AI hardware” and is specifically designed for tasks such as AI inference, machine learning acceleration, and robotics.
Powered by the Nvidia Jetson Orin NX Super platform, Pandora comes in two versions with 8 GB or 16 GB of RAM, offering 117 and 157 AI TOPS, respectively. The device comes pre-installed with a 128GB SSD, Ubuntu Linux, JetPack SDK 6.1.1, and features an active cooling system to maintain performance under load.
Proper selection of ports
Pandora features two USB 3.2 Gen2 Type-A ports, one USB 3.2 Gen2 Type-C/OTC port, and two USB 2.0 Type-A ports for peripherals and legacy devices. It also includes dual Ethernet ports (2 x RJ45, 1G) and an HDMI 2.0 port. For audio, there are both line-out/line-in options via a 3.5mm jack or pin header.
Via the M.2 slot, users can add SSD, Wi-Fi, Bluetooth, and 5G/LTE modules. One of Pandora’s distinguishing features is its support for 3D-printed expansions through customizable casing shells, allowing users to extend the capabilities of their devices. This is a useful option for developers who require adaptable hardware.
Palit suggests several use cases for Pandora. In smart retail, devices can analyze customer demographics and deliver personalized ads in real time through digital signage.
For education, it supports interactive teaching tools using object recognition and text-to-speech technology, and for robotics and automation, it offers extensive I/O options and compatibility with Nvidia Isaac and ROS2. It can also handle demanding generative AI tasks such as natural language processing and content creation.
Pricing and release details for Pandora have not yet been announced, but we should know more soon.
watch on
