Ubuntu for Rockchip RK35XX Devices

FFmpeg with async and zero-copy Rockchip MPP & RGA support

an edge-real-time anchor-free object detector with decent performance

This source code builds openwrt for Rockchip Soc from upstream source code https://github.com/coolsnowwolf/lede

Windows on Arm drivers for RK35xx platforms.

Minimal NixOS running on RK3588/RK3588s based SBC(Orange Pi 5 Plus, Orange Pi 5, Rock 5A, etc)

YoloV5 NPU for the RK3566/68/88

Real-time YOLOv8n UAV detection on the RK3588S NPU | saturates the sensor at 46 FPS | ~140 MB RAM | fully hardware-accelerated (ISP+RGA+NPU) | on-device LLM scene summaries

Run your yolov7 object detection with Rockchip NPU platforms (RK3566, RK3568, RK3588, RK3588S, RV1103, RV1106, RK3562).

Examples to use Rockchip RK3588s RGA2-E/RGA3 platform.

基于u2net网络进行简单修改使其部署到rk3588板子上

rtop, a performance monitor for the Rockchips RK3566/68/88

Entire pipeline for training, converting and exporting a YOLO model into RockChip NPU

Matrix multiplication on the NPU inside RK3588

Good Good Study Day Day Up !!!!

Ubuntu 26.04 (Resolute) for the Orange Pi 5 Pro / RK3588S — mainline kernel with it all working: Mali-G610 GPU + Vulkan (PanVK), RK3588 NPU on all 3 cores (multi-core, 800 MHz DVFS, ~558 inf/s), HW video decode (VA-API), and KDE Plasma. Prebuilt images + Armbian build recipe.

Entire pipeline for running optimized LLM models on the RK3588S, both on the NPU (RKLLM) or the CPU (llama)

Run Yolov8 on Rockchip NPU (from Orange PI 5)

Linux kernel 5.10 (Rockchip upstream) for RK3588.

rtop, a performance monitor for the Rockchips RK3566/68/88