It runs on Rockchip RK3399Pro high-performance AI processor with built-in neural network unit (NPU), supports multiple AI development tools and interfaces, and has rich expansion interfaces and powerful hardware encoding and decoding abilities, which can be applied to AI industry easily.
Rockchip RK3399Pro processor adopts the architecture of dual-core Cortex-A72 and quad-core Cortex-A53 with its frequency up to 1.8GHz, showing an ultra-strong general-purpose computing performance. Quad-core ARM high-end GPU Mali-T860 integrates more bandwidth compression techniques and therefore has an excellent overall performance.
AIO-3399ProC can directly adopt TensorFlow/Caffe/Mxnet general-purpose model and provides AI development tools like model transformation and end-to-side API. It also supports the development interfaces of Android NN API, RKNN cross-platform API and TensorFlow.
It supports various display and output interfaces including HDMI 2.0, MIPI-DSI, eDP and dual LVDS. It also supports dual-screen identical display/dual-screen differential display, showing powerful hardware encoding and decoding abilities. Besides, and supports 4K VP9, 4K 10bits H265/H264, 1080P multi-format (VC-1, MPEG-1/2/4, VP8) video decoding and 1080P (H.264/VP8 format) video encoding.
It supports CAN bus data communications with highly-efficient real-time, farther transmission distance and stronger anti-electromagnetic interference ability. It configures independent external hardware watchdog and makes the device work continuously in the unmanned state which better promotes the system's stability.
With dual MIPI CSI interfaces and in-built dual ISP, supporting single 13 Mpixel or dual 8Mpixel at the maximum. It can achieve the simultaneous input of dual-camera data and supports high-level processing like gesture detection, deep detection and 3D.
With stable and reliable performance, AIO-3399ProC supports multiple operating systems including Android,Linux+QT and Ubuntu.
On-board I2C, SPI, UART, ADC, PWM, GPIO, PCIe, USB3.0, RS232, RS485, I2S (supporting 8-way digital microphone array input ) and other interfaces. It supports the power supply mode of POE+ (802.3 AT, 30W output power ).
AIO-3399ProC can be equipped with an industrial metal case and 10.1-inch IPS HD multi-point touching screen which is integrated into high-performance application PC. It can also be equipped with industrial case independently and embedded into all kinds of smart devices flexibly.
Dual-core Cortex-A72+ Quad-core Cortex-A53 big.LITTLE core CPU architecture,frequency up tp 1.8GHZ
ARM Mali-T860 MP4 Quad-core GPU
Support OpenGL ES 1.1/2.0 /3.0/3.1， OpenVG1.1， OpenCL， DX11
Support AFBC (frame buffer compression)
Buit-in neural network processor NPU,powerful computing performace:
-Suppprt 8bit/16bit operation,comuting performace up tp 3.0TOPS.
-Compuard with using the traditional GPU as the large chip scheme of AI computing unit,
the power consusumption of NPU is merely 1% of that of GPU.
-Load Caffee/Mxnet/TensorFlow models directly.
-Provide AI development tools:Support model fast conversion,support end-to-side API,
support TensorFLow/TF Lite/Caffe/ONNX/Darknet models
-Provide AI application development interface:Support Android NN API,provide RKNN cross-platform API,
Linux support for TensorFlow development.
High-speed eMMC 5.1 (16GB/32GB/64GB128GB).Support TF card expansion
LPDDR3 3GB (NPU 1GB + CPU 2GB)、LPDDR3 6GB (NPU 2GB + CPU 4GB)
10 / 100 / 1000 MbpsEthernet Interface（RJ45）
- 1 x HDMI 2.0 , Support 4K@60HZ output and HDCP 1.4/2.2
- 1 x MIPI-DSI , Support single channel 1080P@60fps output or dual-channel LVDS 1920x1200@60fps output
(AIO-3399ProC defaults to dual LVDS)
- 1 x eDP 1.3 ( 4 lanes with 10.8Gbps )
Support dual-screen identical display/dual-screen differential display
1 x HDMI 2.0 Audio output
1 x I2S For audio input and output (supports 8-way digital microphone array inputs)
1 x Speaker Two-channel speaker（4Ω，10W 8Ω，5W）
1 x Headset output
1 x Mic Audio input
2x MIPI-CSI camera interfaces ( built-in dual-ISP, Maximum support single 13Mpixel or dual 8Mpixel )
3×USB2.0 Hub, 1 x USB3.0, 1xTYPE-C(OTG)
SPI×1, UART×2, Debug×1, RS485×1(CANx1 and RS485 share the same interface), RS232×1,
ADC×1, TPx1, IRx1, I2C, PWM, GPIO
12V DC input voltage
Provide AI development tools: Support model fast conversion, support end-to-side API, support TensorFlow / TF Lite / Caffe / ONNX / Darknet models.
Provide AI application development interface: Support Android NN API, provide RKNN cross-platform API, Linux support for TensorFlow development.
138 mm × 91.3 mm