RK1808K Rockchip AIoT Dual-Core Cortex-A35 Edge Computing SoC with 3.0TOPS NPU H.264 Codec FCCSP-420

RK1808K is a high-performance AI SoC. The RK1808K features a 3.0 TOPS NPU. For edge AI, RK1808K is the ideal choice. The RK1808K supports industrial-grade temperatures. We recommend RK1808K for machine vision. Choosing RK1808K ensures robust AI inference.

RK1808K Product Overview

RK1808K is a high-performance AI SoC from Rockchip designed for AIoT and edge computing applications. Built on a dual-core Cortex-A35 processor architecture with a maximum clock speed of 1.4GHz, the chip integrates a self-developed NPU delivering up to 3.0 TOPS of INT8 inference performance. It supports INT8/INT16/FP16 precision and is compatible with mainstream deep learning frameworks including TensorFlow, Caffe, ONNX, and Darknet.

RK1808K features a hardware VPU supporting 1080p@60fps H.264 decoding and 1080p@30fps H.264 encoding. The chip comes in an FCCSP-420 package (14mm × 14mm) and supports hardware encryption, MIPI CSI camera interface, USB3.0, Ethernet, and multiple serial ports. With an industrial-grade wide temperature range of -40°C to 85°C, RK1808K is well-suited for AI edge computing in demanding environments.

RK1808K Core Features

RK1808K Core Architecture: Dual-core Cortex-A35, up to 1.4GHz

RK1808K NPU Performance: 3.0 TOPS for INT8 / 300 GOPs for INT16 / 100 GFLOPs for FP16

RK1808K AI Framework Support: TensorFlow, Caffe, ONNX, Darknet

RK1808K Video Processing: 1080p@60fps H.264 Decode / 1080p@30fps H.264 Encode

RK1808K Package: FCCSP-420 (14mm × 14mm)

RK1808K Operating Temperature: -40°C ~ 85°C (Industrial Grade)

RK1808K Interfaces: MIPI CSI camera, USB3.0, Ethernet, multiple serial ports, GPIO

RK1808K Security: Hardware encryption engine

RK1808K Applications

RK1808K Machine Vision & AI Inference: Industrial inspection, smart cameras, facial recognition

RK1808K Edge Computing Devices: AI compute cards, intelligent gateways, edge nodes

RK1808K Industrial Automation: Power line inspection, railway monitoring, Industry 4.0 equipment

RK1808K Multimedia Applications: Video codec processing, streaming devices

RK1808K AIoT: Smart home, smart city terminal devices

RK1808K Key Advantages

RK1808K High-Performance NPU (3.0 TOPS): Built-in self-developed NPU delivers powerful AI inference with INT8/INT16/FP16 multi-precision support

RK1808K Industrial-Grade Wide Temperature: -40°C to 85°C operating range for demanding industrial environments

RK1808K Complete AI Ecosystem: Compatible with TensorFlow, Caffe, ONNX, Darknet with RKNN development kit for accelerated deployment

RK1808K Rich I/O Interfaces: MIPI CSI camera, USB3.0, Ethernet, multiple serial ports for diverse peripheral expansion

RK1808K Hardware Video Codec: 1080p@60fps H.264 decoding and encoding offloads CPU

RK1808K Hardware Encryption Engine: Ensures data security and application software copyright protection

RK1808K Low Power & High Efficiency: Cortex-A35 architecture balances performance and power consumption for sustained edge workloads

RK1808K Flexible Deployment: Supports Mini-PCIe compute cards, SOMs and other form factors for rapid productization

Why Choose QIXINWEI

Years of experience in the electronics industry. Trusted by global customers.
Massive In-Stock Inventory – Ready to ship promptly.
BOM Matching Service – One-stop solution, save time.
PCBA Customization – Professional engineering team creates tailor-made solutions based on your needs.
Cost-Effective & Efficient – Better channel, better cost.
A dedicated team makes your procurement smoother.
Contact us for BOM quotes or PCBA inquiries.

FAQ

1. What is RK1808K?
RK1808K is a dedicated AI processor SoC from Rockchip for AIoT and edge computing, available in an FCCSP-420 package. It features a dual-core ARM Cortex-A35 CPU and a built-in NPU (Neural Processing Unit) with peak computing power of 3.0 TOPS, making it Rockchip's first chip designed specifically for AI applications.

2. What are the key specifications of RK1808K?
Dual-core ARM Cortex-A35 processor, up to 1.4GHz; 3.0 TOPS NPU supporting INT8 (3TOPS) / INT16 (300 GOPs) / FP16 (100 GFLOPs) mixed precision; supports 800MHz 32-bit LPDDR2/LPDDR3/DDR3/DDR3L/DDR4 memory; supports TensorFlow, Caffe, ONNX, Darknet, and other mainstream AI frameworks.

3. What package does RK1808K use?
FCCSP-420L package, measuring 14mm × 14mm.

4. What applications is RK1808K suitable for?
Machine vision, AI inference, Industry 4.0, edge computing devices, multimedia applications, intelligent security, and AI algorithm edge deployment. In Mini-PCIe compute card form, it can quickly add AI capabilities to traditional embedded devices.

5. What development tools support RK1808K?
RKNN-Toolkit for NPU model conversion and deployment, supporting model conversion from TensorFlow, Caffe, ONNX, Darknet, and other frameworks. Also provides a complete Linux SDK (based on Debian 10) with integrated RGA, MPP, and RKNN system software libraries.

6. What is the operating temperature range of RK1808K?
RK1808K is a wide-temperature industrial-grade chip with an operating range of -20°C to 85°C. Some module solutions support -40°C to 85°C.

7. What is the NPU computing power of RK1808K?
RK1808K features a dedicated NPU with 3.0 TOPS INT8 peak computing power. Also supports INT16 (300 GOPs) and FP16 (100 GFLOPs) precision. Supports OpenCL/OpenVX.

8. What is the video processing capability of RK1808K?
Supports 1080p@60fps H.264 hardware decoding and 1080p@30fps H.264 hardware encoding. Features a built-in ISP for camera video input; supports 4-lane MIPI-CSI camera interface and MIPI-DSI display interface (up to 1920×1080).

9. What is the difference between RK1808K and RK1808?
RK1808K is the wide-temperature industrial-grade version of RK1808. RK1808 operates from 0°C to 80°C, while RK1808K extends to -20°C to 85°C. Both share the same core specifications: dual-core Cortex-A35 CPU and 3.0 TOPS NPU.

10. What peripheral interfaces does RK1808K support?
Supports USB 3.0/PCIe 2.1, Gigabit Ethernet, 8×UART, 3×SPI, 6×I2C, 11×PWM, 4×SARADC. Audio interfaces include 2-ch & 8-ch I2S, 8-ch PDM with built-in hardware VAD (Voice Activity Detection). Supports 4-lane MIPI-CSI camera interface and 4-lane MIPI-DSI display interface. Supports SPI Nor/NAND Flash, eMMC, and other storage media.