EC-OrinNX
CSR2-N72R3399
AIBOX-OrinNX
CT36L/CT36B
AIBOX-OrinNX
6 TOPS Computing Power
Private Deployment of Large Models
Supports Triple Independent Displays
Supports 4K Encoding and Decoding
Featuring the Rockchip RK3576, the chip adopts a big.LITTLE architecture (4×A72 + 4×A53) with advanced process technology, reaching a clock speed of up to 2.2GHz, providing robust support for high-performance computing and multitasking. Equipped with the Mali-G52 MC3 GPU, offering 145G FLOPS, it enables efficient heterogeneous computing, meeting the demands of graphics-intensive applications.
6 TOPS NPU supports INT4 / INT8 / INT16 / FP16 / BF16 / TF32 operations, enabling dual-core collaboration or independent operation, and supports multitasking and multi-scenario parallelism. It also supports the YOLO (You Only Look Once) real-time object detection algorithm based on deep learning, which offers faster and more real-time performance compared to traditional algorithms, accurately identifying and locating multiple objects in images or videos, empowering AI applications.
Supports private deployment of large-scale parameter models under the Transformer architecture, such as the Gemma series, ChatGLM series, Qwen series, and Phi series of large language models. It also supports traditional network architectures like CNN, RNN, and LSTM, as well as multiple deep learning frameworks including TensorFlow, TensorFlow Lite, PyTorch, and Caffe.
Supports 4K@120fps decoding (H.265 / HEVC, VP9, AVS2, AV1) and 4K@60fps decoding (H.264 / AVC), 4K@60fps encoding (H.265 / HEVC, H.264 / AVC).
Supports multiple display interfaces such as HDMI, MIPI DSI, and LVDS, enabling triple independent displays with 4K@120Hz ultra-clear resolution, meeting diverse display needs across various scenarios.
Equipped with industrial-grade features such as real-time networking, Flexbus, hardware resource isolation, and DSMC, it meets diverse industrial application needs. With support for the RTLinux kernel, it delivers excellent real-time performance and is widely used in industrial application scenarios.
SDK incorporates new features such as AMP, DPDK, RT-Thread, OTA, RT-Linux, and ROS, catering to a broader range of product requirements. It provides accompanying source code, tutorials, technical documentation, and development tools, making development simpler and more convenient.
AIO-3576C | ||
Basic Specifications |
SOC |
Rockchip RK3576 |
CPU |
Octa-core 64-bit processor (4×A72 + 4×A53), up to 2.2GHz |
|
GPU |
G52 MC3 @ 1GHz, supports OpenGL ES 1.1/2.0/3.2, OpenCL 2.0, Vulkan 1.1, embedded high-performance 2D acceleration hardware |
|
NPU |
6 TOPS computing power NPU, support INT4/INT8/INT16/FP16/BF16/TF32 operation, support dual-core collaborative or independent work, support multi-task, multi-scenario parallelism |
|
ISP |
Built-in 16 million pixel ISP, support low-light noise reduction, support RGB-IR sensor, support up to 120dB HDR, AI-ISP to improve low-noise image effect |
|
Codecs |
Decode: 8K@30fps/4K@120fps (H.265/HEVC, VP9, AVS2, AV1), 4K@60fps (H.264/AVC) Encode: 4K@60fps (H.265/HEVC, H.264/AVC) |
|
RAM |
LPDDR5 (4GB/8GB/16GB Optional) |
|
Storage |
eMMC (16GB/32GB/64GB/128GB/256GB Optional) |
|
Storage Expansion |
1 × M.2 M-KEY (Scalable SATA3.0/PCIe NVMe SSD, support 2242/2260/2280), 1 × TF Card |
|
Power |
DC 12V (5.5mm × 2.1mm, support 12V~24V wide voltage input), 1 × DC_IN (4Pin-2.54mm) |
|
Power consumption |
Max: 8.4W(12V/700mA), Normal: 1.2W(12V/100mA), Min: 0.24W(12V/20mA) |
|
OS |
It supports RTLinux kernel and has excellent real-time performance, which is widely used in industrial application scenarios. Support Android14, Linux OS, Buildroot, domestic operating system, provide a safe and stable system environment for product research and production. It has new industrial features such as real-time network, Flexbus, hardware resource isolation, and DSMC to meet the needs of different industrial applications. |
|
AI performance |
Support the privatization deployment of ultra-large-scale parametric models under the Transformer architecture, such as Gemma series, ChatGLM series, Qwen series, Phi series and other large language models. It supports traditional network architectures such as CNN, RNN, and LSTM, and supports the import and export of RKNN models; Support a variety of deep learning frameworks, including TensorFlow, TensorFlow Lite, PyTorch, Caffe, ONNX and Darknet. It also supports the development of custom operators. Support Docker container management technology. It supports the real-time object detection algorithm YOLO (You Only Look Once), which is fast and real-time compared with traditional object detection methods, and can accurately identify and locate multiple target objects in images or videos, powering AI applications. |
|
Size |
126.0mm × 94.4mm × 14.2mm |
|
Weight |
≈91g |
|
Environment |
Operating Temperature: -20℃ ~ 60℃, Storage Temperature: -20℃ ~ 70℃, Storage Humidity: 10%~90%RH(non-condensing) |
|
Interface Specifications |
Internet |
1 × Gigabit Ethernet (RJ45/1000Mbps), supports 2.4GHz/5GHz dual-band WiFi (802.11a/b/g/n/ac), Bluetooth 5.0, expandable 4G LTE (Mini PCIe) |
Video Input |
2 × MIPI-CSI DPHY (1×4Lanes or 2×2Lanes, 30Pin-0.5mm) 1 × MIPI-CSI D/CPHY (DPHY(1×4 Lanes/2×2 Lanes) or CPHY(3Lanes), 30Pin-0.5mm) |
|
Video Output |
1 × HDMI2.1 (4K@120fps, multiplexed with eDP), 1 × eDP (Not pasted by default, multiplexed with HDMI, can be pasted with 30P-0.5mm or 40P-0.5mm two specifications), 1 × MIPI-DSI (2560×1600@60Hz, 30Pin-0.5mm), 1 × LVDS (Dual-channel, double-row pin headers 2×15-30Pin-2mm) |
|
Audio |
1 × MIC (2Pin-2mm), 1 × SPK (4Pin-2mm, 2×10W/6Ω or 2×10W/8Ω), 1 × 3.5mm Audio jack (Support MIC recording, American Standard CTIA) |
|
USB |
1 × USB3.0 (HOST, max: 1.2A), 1 × USB2.0 (HOST, max: 0.5A), 1 × Type-C (OTG), 1 × USB_HOST (4Pin-2mm) |
|
Button |
1 × Reset、1 × Recovery、1 × BOOT |
|
Other interfaces |
1 × RS232 (3Pin-2mm), 1 × RS485 (3Pin-2mm), 1 × CAN (3Pin-2mm), 1 × I2C (4Pin-2mm), 4 × ADC (6Pin-2mm), 1 × Debug (3Pin-2mm), 1 × LCD_PWR (6Pin-2mm), 1 × TP_I2C (6Pin-2mm), 1 × FAN (4Pin-1.25mm) , 1 × Screen voltage jumper (Single-row pin headers, 6Pin-2mm) |