Features: Low power consumption and stable performance.     notes: 1. Compatibility: Can be used with H2, H3, H5, H6, A64 Orange Pi development boards.     2. Provide model conversion tools to support caffe-based convolutional neural network (CNN) model decomposition.     3. The PLAI training tool can directly apply the original image, video, voice, natural language rapid training prototype to the neural network computing stick.          Model: for Orange Pi AI Stick Lite     AI Processor:     Chip Series: Lightspeeur     Chip model: SPR2801s     Energy efficiency: 9.3TOPs/watt     Ultra-low power consumption: 2.8TOPs@300mW     Peak performance: 5.6TOPs@100MHz     Hardware interface: SDIO3.0/eMMC 4.5     Chip package: BGA (7mm*7mm)     Manufacturing process: 28nm     Interface standard: USB2.0, USB3.0 Type-A     Transmission bandwidth: read bandwidth=68.00MB/s, write bandwidth=84.69MB/s     Open source frameworks: Caffe, Tensorflow, PyTorch (Pytorch only supports classification tasks for the time being)     Supported models: VGG, SSD     Operating System: X86 Linux (Ubuntu 16.04) ARM Linux/ ARM Android (ARM v7, ARM v8)     Working voltage: DC 5V 200mA     Operating temperature: 0°C to 40°C     Storage temperature: -20°C to 80°C     Dimensions: 66.5mm x 20.5mm x 10.8mm     Software provided: Plai (based on Pytorch's "full-stack" framework)     Software provided: Open ARM, X86 SDK, built-in 1 to 2 model demos; model training tool Plait.     Material: ABS     colour: White          Package Contents:     1 x for Orange Pi AI Stick Lite          Only the above package content, other products are not included.     Note: Light shooting and different displays may cause the color of the item in the picture a little different from the real thing. The measurement allowed error is +/- 1-3cm.