Core-3568J SMARC

Core-3568J SMARC high-performance AI core board complies with the SMARC 2.0 industry standard and has a compact size of only 82mm x 50mm. It is powe-
red by a Rockchip quad-core 64-bit processor RK3568 ,It equipped with 1 TOPS NPU for supporting lightweight AI application development. With 4K video
decoding capabilities, it boasts abundant interface extensions and supports various video input and output interfaces. It can be expanded with high-speed
network communications such as Gigabit Ethernet, WiFi, Bluetooth, 5G/4G, making it suitable for functional requirements in scenarios such as edge comput-
ing, cloud terminals, IoT, intelligent NVR, industrial automation, and commercial displays.

SMARC 2.0 Industry standard interface

The Core-3568J SMARC is designed in accordance with the SMARC 2.0 industry standard, featuring a 314P MXM3.0 connecting finger interface and a compact size of
only 82mm x 50mm. It is compatible and interoperable with carrier boards and systems that conform to the SMARC standard, significantly enhancing the sys-
tem's flexibility and scalability. This simplifies the design and manufacturing process for high-performance devices, enabling rapid project research and pro-
duction. Additionally, it supports an external watchdog and possesses industrial-grade stability.

Quad-core 64-bit AI Processor RK3568

Core-3568J SMARC with the quad-core 64-bit Cortex-A55 industrial-grade processor RK3568, featuring a 22nm advanced process for low power con-
sumption and high performance, it integrates a dual-core GPU - ARM Mali-G52 2EE graphics processor. This supports mainstream modern graphics
APIs and embeds high-performance 2D acceleration hardware, compatible with OpenGL ES 1.1/2.0/3.2, OpenCL 2.0, and Vulkan 1.1.

1 TOPS computing power for AI applications

Core-3568J SMARC built-in with a 1 TOPS computing power NPU, it supports the development of lightweight AI applications, suitable for edge detection, face
recognition, and other lightweight AI applications.It provides RKNN Toolkit model conversion tool, which supports easy to conversion of mainstream frame-
work models including Caffe, TensorFlow, TFLite, ONNX, PyTorch, Keras, and Darknet.

4K H.265/H.264/VP9 video decoding

It supports 4K@60fps H.265/H.264/VP9 video decoding and 1080P@60fps H.265/H.264 video encoding. It can decode multiple video sources simultaneously,
making it suitable for NVR local display and local decoding in video conferences. With support for 4K resolution and HDR10, it provides an intuitive visual experience and ex-
quisite display effects.

Strong network communication capability

With 1 1000Mbps Ethernet,It can expand WiFi/Bluetooth modules through SDIO/PCIe interfaces, and support the expansion of 5G/4G LTE wireless
communication, enabling higher-speed network communication.

Abundant interface options

This core board provides a rich array of expansion interfaces such as PCIe3.0、PCIe2.0、SATA3.0、 HDMI2.0、MIPI-CSI、MIPI-DSI、LVDS、USB3.0、
USB2.0、SDMMC、SPI、UART、I2C、I2S、PWM、ADC、CAN、GPIO to meet peripheral expansion needs for various applications.

Abundant resources

We offer SDKs, tutorials, technical documentation, and development tools to streamline and improve the development process.

A wide range of applications

It is widely used in edge computing, cloud terminals, the Internet of Things, intelligent NVRs, industrial automation, and commercial displays, and more.

cloud terminals
edge computing
the Internet of Things
industrial automation
commercial displays
intelligent NVRs

Specifications

Specifications
Basic Specificat
SOC

RK3568B2

CPU

Quad-core 64-bit Cortex-A55 processor, 22nm lithography process, up to 2.0GHz

GPU

ARM G52 2EE graphics processor, supports OpenGL ES 1.1/2.0/3.2, OpenCL 2.0, Vulkan 1.1, embedded high-performance 2D acceleration hardware

NPU

1Tops@INT8, RKNN NPU AI accelerators, supports one-click conversion of Caffe/ TensorFlow/TFLite/ONNX/PyTorch/Keras/Darknet architecture models

VPU

4K@60fps H.265/H.264/VP9 video decoding, 1080P@60fps H.265/H.264 video encoding

RAM

2GB LPDDR4 (4GB/8GB optional)

Storage

32GB eMMC (64GB/128GB optional)

Power

5V (voltage tolerance ±5%)

System

Ubuntu, Android

Interface

SMARC standard (314Pin MXM3.0 interface)

Size

82 mm × 50 mm

Weight

≈19g

Environment

Operating Temperature: -20℃~60℃, storage temperature: -20℃~70℃, Storage Humidity: 10%~90%RH (non-condensing)

Ethernet

1 × Gigabit Ethernet (1000Mbps)

Interface Specif
WiFi

The "WiFi + Bluetooth" module (2.4GHz/5GHz dual-band WiFi6, Bluetooth 5.0) can be expanded through the SDIO interface, supporting the expansion of 5G/4G LTE wireless networks

Video output

1 × HDMI2.0 (4K@60fps) 2 × MIPI-DSI (4Lanes, support single-channel 1920×1080@60fps output or dual-channel 2560×1440@60fps output) 1 × LVDS (single channel 1280×800@60fps or dual channel 1920 ×1080@60fps, one channel multiplexed with MIPI-DSI0 channel)

Camera

1 × MIPI-CSI (1 × 4 Lanes or 2 × 2 Lanes)

PCIE

1 × PCIe 3.0(2 Lanes) 1 × PCIe 2.1(1 Lane)

SATA

3 × SATA3.0 (support eSATA)

USB

1 × USB3.0 HOST, 1 × USB3.0 OTG, 2 × USB2.0 HOST

Other

3 × SDMMC, 4 × SPI, 10 × UART, 6 × I2C, 2 × I2S/PCM, 16 × PWM, 5 × ADC, 3 × CAN, GPIOs