ICORE-3572Q38 AI SoM

ICORE-3572Q38 AI SoM

$0.00
Sale price  $0.00 Regular price 
Skip to product information
ICORE-3572Q38 AI SoM

ICORE-3572Q38 AI SoM

$0.00
Sale price  $0.00 Regular price 
Shipping calculated at checkout.

Compact 38 × 38 mm RK3572 AI SoM with 4 TOPS NPU, 8K video decoding, and rich interfaces for AI vision and edge applications

  • American Express
  • Apple Pay
  • Google Pay
  • Mastercard
  • PayPal
  • Shop Pay
  • Union Pay
  • Visa
ICORE-3572Q38

ICORE-3572Q38

ICORE-3572Q38 features Rockchip RK3572 octa-core 64-bit AIoT processor, built on an advanced 8nm process with excellent low-power characteristics. It is equipped with 4 TOPS NPU and an integrated Mali-G310 GPU supporting OpenGL ES 1.1/2.0/3.2 and Vulkan 1.4 graphics APIs. It integrates 12M ISP image processing unit and supports 8K high-definition video decoding, enabling dual-display output at 4K@60fps + 2K@60fps. The board adopts a BTB connector with a compact size of only 38mm × 38mm, offering rich peripheral expansion capabilities.

ICORE-3572Q38
Powered by RK3572 AI Processor

Powered by RK3572 AI Processor

ICORE-3572Q38 features Rockchip RK3572 octa-core 64-bit AIoT processor, built on an advanced 8nm process with a maximum frequency of 2.2GHz, delivering excellent low-power performance. It integrates a Mali-G310 GPU supporting OpenGL ES 1.1/2.0/3.2 and Vulkan 1.4.

8K Video Decoding with Dual-Display Output

8K Video Decoding with Dual-Display Output

It supports 8K@30fps and 4K@120fps high-frame-rate video decoding, as well as 4K@40fps video encoding. It integrates a 12MP ISP with hardware accelerators including Bayer-3DNR, HDR-Merge, DPCC, LDC, and 3A. It supports up to 5 camera inputs (2x MIPI CSI DPHY (1x4 Lanes + 2x2 Lanes) + 1x DVP), and enables dual-display output at 4K@60fps + 2K@60fps.

Rich AIoT Expansion Interfaces

Rich AIoT Expansion Interfaces

It offers extensive expansion interfaces including PCIe2.1, SATA3.1, USB3.0, MIPI-CSI, DVP, MIPI-DSI, PWM, SPI, I3C, UART, and CAN, fully accommodating diverse peripheral connectivity for AIoT scenarios — including display, image capture, audio, and sensors — to meet the varied expansion and development needs of IoT terminals.

Application Scenario

Widely suitable for edge computing, tablets, mobile terminals, AI cameras, intelligent surveillance, NAS, retail devices, digital signage, and other products and fields.

Edge Computing
Edge Computing
Tablet PC
Tablet PC
Mobile Devices
Mobile Devices
Smart Surveillance
Smart Surveillance
Retail Equipment
Retail Equipment
Digital Signage
Digital Signage

Interfaces

Interfaces

Specifications

ICORE-3572Q38
Basic Specifications SOC

Rockchip RK3572

CPU

Octa-core 64-bit processor (2×A73 + 6×A53) with NEON coprocessor, up to 2.2GHz

GPU

ARM Mali G310V2 MC1 GPU, supports OpenGL ES 1.1/2.0/3.2, OpenCL 3.0, Vulkan 1.4, with built-in high-performance 2D acceleration hardware

ISP

Built-in 12MP ISP, integrated hardware accelerators: Bayer-3DNR, HDR-Merge, DPCC, LDC, 3A, LSC, CAC, de-haze enhancement

NPU

4 TOPS (Sparse INT8). Supports INT4/INT8/INT16/FP4/FP8/FP16/BF16 mixed computation and W4A16 asymmetric MAC operation

Codec

Video Decode: 8K@30fps/4K@120fps (H.265/HEVC, VP9, AVS2, AV1), 4K@60fps (H.264/AVC)

Video Encode: 4K@40fps (H.265/HEVC, H.264/AVC)

Image Codec: 4K@60fps MJPG

RAM

LPDDR4/LPDDR4X (2GB/4GB/8GB/16GB optional)

Storage

eMMC (16GB/32GB/64GB/128GB optional)

OS

Debian, Ubuntu

Software support

Supports YOLO object-detection algorithms

Supports private deployment of large-parameter Transformer-based LLMs, including Gemma, ChatGLM, Qwen and Phi series

Supports classic network architectures such as CNN, RNN and LSTM; enables RKNN model import & export; compatible with TensorFlow, TensorFlow Lite, PyTorch, Caffe, ONNX, MXNet deep-learning frameworks

Supports custom operator development and Docker container-based management

Power

5V (voltage tolerance ±5%)

Power consumption

Normal: 1W(5V/200mA), Max: 3.5W(5V/700mA), Min(Sleep): 0.225W(5V/45mA)

Interface type

BTB (2 × BTB Socket (100Pin))

Dimension

38.0mm × 38.0mm × 5.5mm

Weight

≈9g

Environment

Operating Temperature: -20℃~60℃, Storage Temperature: -20℃~70℃, Operating Humidity: 10%~90%RH (No condensation)

Interface Specifications Internet

2 × GMAC (RMII/RGMII for external Ethernet PHY; 10/100/1000 Mbps)Extend Wi-Fi, BT, 4G, 5G via USB/PCIe/SDIO/UART

SATA/PCIe

1 × PCIe 2.1/SATA 3.1/USB 3.0 combo interface (select 2 out of 3 functions)

1 × PCIe 2.1/SATA 3.1 combo interface (select 1 out of 2 functions)

Video input

2 × MIPI-CSI DPHY (MIPI V1.2; 1 × 4 Lanes or 2 × 2 Lanes)

1 × DVP (BT.601/BT.656 & BT.1120; BT.656/BT.1120 supports 2-/4-channel multiplexed byte-interleaved format, double-edge sampling, 8/10/12/16-bit)

Video output

1 × MIPI-DSI DPHY 1.2 (1×4Lanes or 2×2Lanes, 2.5Gbps/Lane)

Audio input

3 × SAI (4R), 2 × SAI (1R), 1 × SAI (4R), supports I2S/TDM/PCM audio protocols, sampling rate up to 192 KHz

Audio output

3 × SAI (4T), 2 × SAI (1T), supports I2S/TDM/PCM audio protocols, sampling rate up to 192 KHz

USB

2 × USB3.0 DRD/USB2.0

SDIO

2 × SDIO3.0

PWM

16 × PWM

SPI

5 × SPI (master/slave configurable by software)

I2C

10 × I2C (7-bit / 10-bit address; up to 100kbps standard, 400kbps fast mode)

I3C

1 × I3C (I2C-compliant, SDR mode, supports up to 11 devices)

UART

12 × UART (supports auto flow-control mode and RS485 function)

CAN

4 × CAN (supports 8192-bit receive FIFO)

Customization

Firefly team provides hardware, software, turnkey and OEM/ODM customization services.

Firefly is dedicated to delivering hardware-software-integrated intelligent solutions across diverse industries. Firefly is dedicated to delivering hardware-software-integrated intelligent solutions across diverse industries.

Contact Us

Onsite SupportOnsite Support

Email SupportEmail Support

Remote Video SupportRemote Video Support

Real-time SupportReal-time Support