RK3576 Rockchip Octa-Core Cortex-A72/A53 AIoT Processor Mali-G52 GPU 6TOPS NPU 4K Codec Dual GbE BGA

Property:
Specification
Product Type:
Octa-Core Cortex-A72/A53 AIoT Processor
Brand:
Rockchip
CPU:
Quad Cortex-A72 2.2 GHz + Quad Cortex-A53 1.8 GHz
GPU:
Arm Mali-G52
NPU:
6.0 TOPS
Package:
BGA
Display Interfaces:
HDMI 2.0, MIPI DSI, eDP
Video Codec:
4K@60fps H.264/H.265/VP9
Ethernet:
Dual Gigabit Ethernet MAC
Temperature:
0°C to 80°C

RK3576 is a powerful octa-core AIoT processor. The RK3576 features a 6.0 TOPS NPU. For edge AI, RK3576 is the top choice. The RK3576 supports dual GbE. We recommend RK3576 for industrial HMI. Choosing RK3576 ensures robust performance.

RK3576 Product Overview

RK3576 is a high-performance octa-core AIoT processor from Rockchip, featuring a big.LITTLE architecture with quad-core Cortex-A72 (up to 2.2 GHz) and quad-core Cortex-A53 (up to 1.8 GHz) to deliver robust general-purpose computing for the edge. It integrates a 6.0 TOPS NPU that supports mainstream frameworks such as TensorFlow, PyTorch, and ONNX, easily handling complex AI tasks like image recognition and voice processing. The Mali-G52 GPU supports OpenGL ES 3.2, Vulkan 1.1, and OpenCL 2.0 for smooth 4K UI and 3D graphics. The video engine provides 4K@60fps H.264/H.265/VP9 hardware encoding and decoding, with HDMI 2.0, MIPI DSI, and eDP display interfaces supporting dual independent displays. For networking, it integrates dual Gigabit Ethernet MACs, along with USB 3.0, PCIe 3.0, and a rich set of UART/SPI/I2C peripherals. Memory support includes DDR4/LPDDR4/LPDDR4x up to 8 GB; boot options cover eMMC 5.1, SD card, and SPI Flash. With its powerful AI, exceptional multimedia, and dual-GbE connectivity, the RK3576 is widely used in AIoT edge gateways, smart NVRs, machine vision, industrial HMIs, and smart retail applications.


RK3576 Core Features

RK3576 CPU: Quad Cortex-A72 (2.2 GHz) + Quad Cortex-A53 (1.8 GHz), big.LITTLE

RK3576 GPU: Arm Mali-G52, OpenGL ES 3.2 / Vulkan 1.1 / OpenCL 2.0

RK3576 NPU: 6.0 TOPS, supports mainstream AI frameworks

RK3576 Video: 4K@60fps H.264/H.265/VP9 encode & decode

RK3576 Display: HDMI 2.0, MIPI DSI, eDP, dual independent displays

RK3576 Memory: Supports DDR4/LPDDR4/LPDDR4x, up to 8 GB

RK3576 Storage Boot: eMMC 5.1, SD/SDIO 3.0, SPI Flash

RK3576 High-Speed Interfaces: USB 3.0 ×1, USB 2.0 ×1, dual Gigabit Ethernet MAC

RK3576 Expansion: PCIe 3.0 ×1, 10× UART, 6× I2C, 5× SPI, 8× PWM

RK3576 Package: BGA

RK3576 Temperature Range: 0°C to 80°C


RK3576 Applications

RK3576 AIoT Edge Gateways: 6TOPS NPU for local inference, dual GbE for internal/external network isolation and high-speed forwarding

RK3576 Smart NVR & Security: Multi-channel video decoding with AI detection, dual Ethernet for multi-subnet access

RK3576 Machine Vision & Industrial Cameras: NPU-accelerated image recognition and measurement, USB 3.0/GbE image transmission

RK3576 Industrial HMI & Panel PCs: Rich industrial bus interfaces for stable human-machine interfaces and device connectivity

RK3576 Smart Retail & Self-Service: Integrated multi-display, AI recognition, and remote management


RK3576 Key Advantages

RK3576 6TOPS High-Performance NPU: Powerful neural network accelerator for real-time inference of complex deep learning models

RK3576 Octa-Core CPU: A72+A53 big.LITTLE design balances performance and efficiency for smooth Linux/Android operation

RK3576 4K@60fps Codec: Hardware encode/decode ensures ultra-HD video processing for multimedia and security needs

RK3576 Dual Gigabit Ethernet: Two independent MACs easily enable network isolation, link aggregation, or multi-subnet connectivity

RK3576 Triple Display Outputs: HDMI + MIPI + eDP flexibly adapt to various screens and dual-display scenarios

RK3576 Large Memory & High-Speed I/O: Up to 8 GB memory with USB 3.0 and PCIe 3.0 for worry-free expansion

RK3576 Mature Software Ecosystem: Android/Linux dual-system support with official SDK and extensive community resources


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 RK3576?
RK3576 is a high-performance, low-power AIoT application processor from Rockchip, built on an 8nm process. It features an octa-core 64-bit CPU (4×Cortex-A72 + 4×Cortex-A53), an ARM Mali-G52 MC3 GPU, and a 6 TOPS NPU, designed for AIoT, edge computing, ARM PCs, and multimedia applications.

2. What are the key specifications of RK3576?
Octa-core 64-bit CPU: 4×Cortex-A72 @ 2.2GHz + 4×Cortex-A53 @ 1.8GHz; ARM Mali-G52 MC3 GPU supporting OpenGL ES 1.1/2.0/3.2, OpenCL 2.0, and Vulkan 1.1; 6 TOPS NPU supporting INT4/INT8/INT16/FP16/BF16/TF32 mixed precision; supports up to 16GB LPDDR4/LPDDR4X/LPDDR5 memory.

3. What package does RK3576 use?
FCCSP-698L package, measuring 16.1mm × 17.2mm. It is significantly smaller than RK3588's 23mm × 23mm package. RK3576J is the industrial-grade version with an operating temperature range of -40°C to +85°C; RK3576 is the wide-temperature grade with a range of 0°C to +80°C.

4. What applications is RK3576 suitable for?
Edge computing, AI large language model private deployment (Gemma, ChatGLM, Qwen, etc.), industrial automation control, smart digital signage, cloud terminals, industrial control hosts, automotive electronics, facial recognition devices, smart home, and ARM PCs.

5. What development tools support RK3576?
Rockchip's Linux SDK (supporting Ubuntu/Debian/Buildroot, etc.) and Android SDK. For NPU deployment, RKNN-Toolkit2 provides model conversion, inference, and performance evaluation, supporting model conversion from TensorFlow, PyTorch, Caffe, ONNX, and other mainstream AI frameworks; RKNN SDK (with C++/Python APIs) is used for on-device inference deployment.

6. What is the operating temperature range of RK3576?
RK3576 offers two temperature grades:

  • RK3576 (wide-temperature grade) : 0°C to +80°C

  • RK3576J (industrial grade) : -40°C to +85°C

7. What is the NPU computing power of RK3576?
RK3576 features Rockchip's proprietary NPU with 6 TOPS INT8 computing power. Supports INT4/INT8/INT16/FP16/BF16/TF32 mixed precision. Supports private deployment of large language models under Transformer architecture, including Gemma, ChatGLM, Qwen, and others.

8. What is the video processing capability of RK3576?
Supports 8K@30fps or 4K@120fps video decoding (H.265/VP9/AVS2/AV1), and 4K@60fps video encoding (H.265/H.264). Features a 16MP ISP with support for up to 120dB HDR. Supports HDMI 2.1 (4K@120fps) / eDP 1.3 / DP 1.4 / MIPI DSI display interfaces, with support for triple-display with different sources.

9. What are the alternative models for RK3576?
Alternatives in the same series include:

RK3588: Flagship model with stronger CPU performance (A76+A55), suitable for high-performance computing and ultra-HD video processing

RK3576J: Industrial-grade version with -40°C to +85°C wide-temperature range

RK3568: Mid-range quad-core Cortex-A55 with 1 TOPS NPU

RK3576 offers approximately 70% of RK3588's performance at about 30% of the cost, delivering exceptional cost-effectiveness for AIoT, industrial, and medium-visual-computing applications.

10. What peripheral interfaces does RK3576 support?
Supports dual Gigabit Ethernet (RGMII×2), PCIe 2.1, USB 3.0, SATA 3.0, dual CAN FD, UFS 2.0, eMMC 5.1, SPI Nor/Nand. Display interfaces include HDMI, DP, eDP, MIPI-DSI, RGB, and EBC. Camera interfaces support MIPI CSI-2 (CDPHY=1×4-lane, DPHY=2×4-lane/4×2-lane) and DVP. Low-speed interfaces include UART, SPI, PWM, I2C, I3C, and SAR-ADC.