NPU: Support 8-bit/16-bit Inference, Support TensorFlow/Caffe/TF Lite Model;
Soc:Rockchip RK3399Pro(NPU built-in); CPU: Dual Cortex-A72 + Quad Cortex-A53, 64-bit CPU, Frequency up to 1.8GHz;
Memory: 3GBLPDDR3;EMMC: 16GB eMMC; Support dual system Android 8.0 and Fedora 28;
Support: Gigabit Ethernet and wifi ;Power supply: 12V-2A; PCB Dimension: 145mm*106mm
Computing architecture and NPU dedicated to AI acceleration, which can be used in build Deep Learning, Robots, Automobile, Face Recognition and Smart City.