Wholesales OEM ODM factory Rockchip RK3399Pro 96Boards SOM Core Board TB-96AI NPU MPU With SDK
- Shipping fee:
- Depends on the order quantity.
- Lead Time:
- 3 day(s) after payment received
Customized logo (Min. Order: 1000 Pieces)Customized packaging (Min. Order: 1000 Pieces)
MoreGraphic customization (Min. Order: 1000 Pieces) Less
Display Port, Ethernet, PCI-Express X4, USB 2.0, USB 3.0, WiFi, Bluetooth
RK3399 Pro, RK3399 PRO
- Products Status:
- Chipset Manufacturer:
- Socket Type:
- Form Factor:
- Memory Type:
- Graphics Card Type:
- Brand Name:
- Place of Origin:
Guangdong, China (Mainland)
RK3399Pro SOM Core Board
- Deliver Time:
3 Working Days
- Product Name:
RK3399Pro SOM Core Board
Packaging & Delivery
- Selling Units:
- Single item
- Single package size:
- 11.8X7.8X3.3 cm
- Single gross weight:
- 1.5 kg
- Lead Time :
Quantity(Piece) 1 - 1 >1 Est. Time(days) 3 To be negotiated
RK3399Pro Core Board
- lThe world's first core board in line with 96Boards System-on-Module (SOM) specifications;
- lDual-core ARM Cortex-A72+ four core ARM Cortex-A53, with superb general computing capability;
- lIntegrated neural network processor NPU,AI calculation force up to 3.0Tops, compatible with Caffe/Mxnet/TensorFlow model;
- l85mmx50mm Ultra-small size, flexible deployment, Panasonic high-speed onboard connector, stable and reliable
- lSystem Support Android/linux/ubuntu
Product Specification V1.0
1. Product overview
On April 1, 2019, Linaro officially released the 96 Boards System-on-Moudle (SoM) specification V1.0 in Bangkok, Thailand. On this basis, two core modules TB-96AI and TB-96AIoT, developed by Xiamen Beiqi Technology Co., Ltd., which conform to the 96 Boards SoM specification, were solemnly launched. TB-96AI uses RK3399Pro as the main control chip and TB-96AIoT uses RK1808 as the main control chip.
TB-96AI is a powerful core board for artificial intelligence. Carrier Board developed by Xiamen Beiqi
Technology Co., Ltd. can form a complete development board or evaluation board; and the board customized by customers according to actual needs can directly form the industry application motherboard, which can meet industrial automation, UAV, image detection, face recognition, edge computing gateway, cluster server, Intelligent Quotient display, automatic driving, medicine. Application needs of market segments such as health care equipment, robots and intelligent retail.
The following features are quoted from RockChip. If you have any questions, please contact BEIQICLOUD for more technical support.
1.2.1 Six-core 64-bit processor, superior general-purpose computing power
l Dual-core ARM Cortex-A72 MPcore processor and quad-core ARM Cortex-A53 MPcore processor are high-performance, low-power and cache application processors.
l Two CPU clusters. Big cluster with dual-coreCortex-A72 is optimized for high-performance and little cluster with quad-core Cortex-A53 is optimized for low power.
l Full implementation of the ARM architecture v8-A instruction set, ARM Neon Advanced SIMD (single
instruction, multiple data) support for accelerating media and signal processing See RK3399Pro datasheet for more features.
1.2.2 Built-in Neural NetworkProcessor NPU, Ultra High AI Computing Power
Supporting 8 bit/16 bit operation, AI computing power up to 3.0 TOPs (INT8 Inference); (300 GOPs for INT16, 100 GFLOPs for FP16 )
Real time rate
accuracy rate WER
l Full load calculation is strong and light load operation power consumption is low.
l Compatible with Caffe/Mxnet/TensorFlow model, it can support multiple frameworks, support mainstream layer types, and add custom layer easily.
l Provide easy-to-use development tools, PC can complete model conversion, performance prediction, accuracyverification.
l Provide AI application development interface: support Android NN API, RKNN cross-platform API, Linux support TensorFlow development;
1.2.3 Powerful Multimedia Processing Performance
l Integrated quad-core ARM Mali-T860MP4 GPU, support OpenGL ES1.1/2.0/3.0, OpenCL1.2, Directx11.1, etc., with more bandwidth compression technology
l Strong hardware codec capability
Ø Support 4K VP9 and 4K 10bits H265/H264 video decoding up to 60fps
Ø Support 1080P multi-format video decoding (VC-1, MPEG-1/2/4, VP8)
Ø Support 1080P video encoding, support H.264, VP8 format
1.2.4 Multiple video input and output interfaces
l Dual camera interface: two MIPI-CSI input interfaces with two ISP image processors
l Display output interface: Embed two VOPs, support dual-screen simultaneous/dual-screen display, and can choose to output from the following display interface.
Ø DP×1(Support progressive/interlaced, support RGB/yuv420/yuv422/yuv44format)
Ø HDMI×1(Support 480p/480i/576p/576i/720p/1080p/1080i/4k, support RGB format)
1.2.5 Rich expansion interface
A rich set of expansion interfaces for users to choose to support I2C, SPI, UART, ADC, PWM, GPIO, PCIe, USB3.0, I2S, etc.
l USB3.0×1,According to the RK3399Pro design, the NPU needs to be mounted on the USB3.0, so the USB3.0 needs to be connected back to the NPU. If you need to extend the USB3.0 interface, you need to plug in the HUB.
l CPU Debug UART×1,NPU Debug UART×1;
l GPIO,For detailed GPIO definition, please refer to interface definition.;
l ADC×3,One for buttons, one for headset microphone detection, and one for user-definable use;
1.2.6 High-speed on-boardconnector for more stability and reliability
l 4 Panasonic high-speed onboard connectors for higher speed signal stability
The core board can be fixed by 4 screw posts for various working environments.
1.2.7 Ultra-high integration, ultra-small size
l The core board integrates RK3399Pro, CPU DDR, NPU DDR, eMMC, power management module, and Ethernet PHY chip. It has high integration, greatly reduces the design difficulty of the application backplane, and helps enterprises to quickly develop mass production specific application products. .
l The design size is only 85mm × 50mm, which can be more easily and flexibly deployed on various types of application boards.
1.2.8 Support for multiple operating systems
l Support Android,Linux, Ubuntu
l Support U disk upgrade through USB interface
Dual-core Cortex-A72 up to 1.8GHz
Quad-core Cortex-A53 up to 1.4GHz
ARM® Mali-T860 MP4 Quad-core GPU
Ø Support OpenGL ES1.1/2.0/3.0/3.1, OpenVG1.1, OpenCL, DX11
Ø Support AFBC (frame buffer compression)
Ø Support 8bit/16bit computing, AI computing power up to 3.0TOPs
Ø Full load computing power, low load operation power consumption is low
Ø Compatible with Caffe/Mxnet/TensorFlow model, support multi- class framework, support mainstream layer type, easy to add custom layer
Ø Provides easy-to-use development tools, PC-based model conversion, performance estimation, and accuracy verification
Ø Provide AI application development interface: support Android NN API, provide RKNN cross-platform API, Linux support TensorFlow
Ø Support 4K VP9 and 4K 10bits H265/H264 video decoding, up to 60fps
Ø 1080P multi-format video decoding (WMV, MPEG-1/2/4, VP8)
Ø 1080P video encoding, support H.264, VP8 format
Ø Video post processor: de-interlacing, denoising, edge/detail/color optimization
Optional configuration with the following two options:
Ø 3GB LPDRR3(CPU 2GB + NPU 1GB);
Ø 8GB LPDDR3(CPU 4GB + NPU 4GB);
Optional configuration with the following options:
Ø 16GB eMMC
Ø 32GB eMMC
Ø 64GB eMMC
Ø 128GB eMMC
Built-in Gigabit Ethernet PHY chip, 10/100/1000Mbps adaptive
MIPI-CSI×2,Dual camera interface (built-in dual hardware ISP, up to
single 13Mpixel or dual 8Mpixel)
Embed two VOPs, support dual-screen simultaneous/dual-screen display, and can choose to output from the following display interface.
Ø HDMI × 1 ( Support 480p/480i/576p/576i/720p/1080p/1080i/4k, support RGB format)
Support user extended use
Ø HDMI interface audio output;
Ø DP interface audio output;
Ø USB3.0× 1 (according to RK3399Pro design, NPU needs to be mounted on USB3.0, so USB3.0 needs to connect back to NPU, if you need to expand USB3.0 interface, you need external HUB);
Ø USB2.0×2, HOST;
Ø SDMMC(TF Card)×1;
Ø UART×3,One of the CPU Debug UARTs, one NPU Debug UART;
Ø GPIO,For detailed GPIO definitions, please refer to the interface
Ø ADC×3,One for buttons, one for headset microphone detection, and one for user-definable use;
Android8.1;Linux version: fedora 2.8, kernel 4.4
Ø Support 8bit/16bit computing, AI computing power up to 3.0TOPs;
Ø Full load computing power, low load operation power consumption is low;
Ø Compatible with Caffe/Mxnet/TensorFlow model, support multi- class framework, support mainstream layer type, easy to add custom layer;
Ø Provide easy-to-use development tools, PC-side model conversion, performance estimation, and accuracy verification;
Ø Provide AI application development interface: support Android NN
API, provide RKNN cross-platform API, Linux support TensorFlow development;
4 Panasonic 100PIN high speed connectors, type AXK6S00437YG (PIN
3. Structure Size
4. TB-96AI+Carrierboard Guide for use
The following figure shows the use of TB-96AI RK3399Pro SOM on the Carrier board.
The following figure shows the interfaces on the board that can be provided to TB-96AI RK3399Pro SOM
For more detail, please send an inquriy as below:
To ensure efficient and qualified production, our company fully implements ISO9001-2008 quality management system. The products have won international certification such as
Meet Us Next
The supplier’s transactions conducted via Alibaba.com in the past 6 months.