The RK3576 development board features a powerful quad-core Cortex-A72 and quad-core Cortex-A53 CPU, Mali-G52 GPU with advanced graphics support, and a 6.0T NPU for AI processing. It offers 4/8GB LPDDR4X memory, external EMMC/UFS storage options, dual GMAC Ethernet ports, and integrated 2.4G/5G WiFi and Bluetooth. With USB 3.0 OTG and Type-C power input, it supports 4K displays via MIPI DSI and HDMI 2.1. Equipped with MIPI-CSI camera interfaces and a 40Pin GPIO for expansion, it includes an OLED display and a Recovery button. Powered by Linux, this compact board measures 87mm×57mm.
High-Performance CPU and GPU: Equipped with a quad-core Cortex-A72 and quad-core Cortex-A53 CPU along with a Mali-G52 GPU, this board delivers powerful processing capabilities for demanding applications.
Advanced AI Processing: Features a 6.0T NPU that supports INT4, INT8, INT16, and FP16 mixed operations, making it suitable for AI and machine learning tasks.
4K Video Capabilities: Offers 4K@120fps decoding and 4K@60fps encoding, providing robust support for high-resolution video processing.
Flexible Connectivity: With dual GMAC Ethernet ports, onboard 2.4G/5G WiFi and Bluetooth, and multiple USB interfaces, this board provides a range of options for wired and wireless connections.
Compact Design with External Storage: The compact 87mm×57mm form factor combined with support for external EMMC/UFS storage makes it ideal for space-constrained applications while offering expandable storage options.
x1 Toybrick TB-RK3576D SBC
| Specification | Details |
|---|---|
| Main Controller Chip | RK3576 |
| CPU | Quad-core Cortex-A72 and quad-core Cortex-A53 |
| GPU | Mali-G52 GPU |
| GPU Features | Supports OpenGL ES1.1, 2.0, 3.2 / OpenCL 2.0 / Vulkan 1.1 |
| GPU Features | Integrated MMU with dedicated 2D hardware engine |
| VPU | Decode: 4K@120fps H.264/AV1/VP9/AVS2, 4K/60fps H265 |
| VPU | Encode: 4K@60fps H.264/H.265 |
| VPU | Codec: 4K@30fps MJPG |
| NPU | Supports 6.0T computing power, INT4, INT8, INT16, FP16 mixed operations |
| NPU | Supports model conversion based on TensorFlow / MXNet / PyTorch / Caffe frameworks |
| Memory | 4/8GB LPDDR4X |
| Storage | External EMMC/UFS small board (optional) |
| Ethernet | 2×GMAC(10/100/1000M) |
| Wireless Network | Onboard WIFI module: Supports 2.4G/5G WiFi, BT |
| USB | 2 USB3.0 OTG interfaces |
| USB | 1 TypeC interface (power in) |
| Display Interfaces | 1×MIPI DSI interface, supports 2560×1600@60fps |
| Display Interfaces | 1×HDMI 2.1 interface, supports 4K@120fps |
| Camera Interface | 2 MIPI-CSI 4lane camera interfaces (Officially supported camera models: IMX415, OV50C40) |
| Debug | 1 debug serial port (via 40pin GPIO pin37 and pin40) |
| Expansion Interface | 40Pin GPIO, supports UART, SPI, I2C, PWM, etc. |
| TF card | 1 TF card slot, supports SD3.0 |
| Buttons | 1 Recovery button |
| Power Input | Type C interface, only power input (supports PD2.0/3.0) |
| System | Linux |
| Core Board Size | 87mm×57mm |
